# Will Dzierson **Product, design & engineering.** Product and design leader with 25+ years of experience. A hands-on builder working across AI, healthcare, and human-centered systems. - **Current role:** VP, Product & Design at [InsideTracker](https://www.insidetracker.com/) - **Experience:** 25+ years in design and technology - **Location:** Saratoga Springs, New York - **Email:** [will@dzierson.com](mailto:will@dzierson.com) - **Elsewhere:** [LinkedIn](https://www.linkedin.com/in/will-dzierson-1081963/) · [Medium](https://medium.com/@willdzierson) I conceive, design, build, and ship. 25+ years connecting what’s possible with what people actually need. ## About I sit at the intersection of product, design and engineering — I love to conceive, to design, and to build. At Google, I led design on mobile experiences that [reshaped how millions of users](https://www.youtube.com/watch?v=JKxzX3p1iRs) interact with information on their phones. Since then I've shipped software for healthcare, AI, and consumer products at Salesforce, Included Health, Dexterity Robotics, and a dozen others. I'm drawn to the hard problems: building AI interfaces that feel intuitive, designing systems that scale, and bridging the gap between what's technically possible and what's genuinely useful. The most useful place to work is often between disciplines. I connect product strategy, research, design, and engineering—then stay with the work through implementation. Today, I lead Product & Design at InsideTracker. I still build. ## Specializations - Product leadership and building design and research teams - Product strategy - UX research and interaction design - Design engineering and TypeScript prototyping - AI products: RAG, conversational, voice, multimodal, and agentic interfaces - Healthcare and digital health - Mobile design - Founding and 0-to-1 products ## How I work ### Product leadership: Lead through the work. At Grand Rounds, my work progressed from lead and principal design roles to Design Director, growing and coaching the Design & Research teams. My current VP role continues that player-coach approach. (Team development · Product direction · Cross-functional delivery) ### Design & engineering: Make the decisions testable. At Included Health, I built conversational prototypes and worked with clinical, product, and engineering partners. A comparative study of conventional, chat, and voice interfaces informed subsequent AI interaction patterns. [Explore the multimodal work](https://dzierson.com/project/cc5b34e1-e0a7-4a59-b3b6-c9e814585c34) ### Founding: Own the problem and the product. At Noodle, I co-founded the company, helped raise angel funding, led design, and co-led engineering with contract engineers. The company dissolved; the experience remains part of how I evaluate product bets and execution. [Explore Noodle](https://dzierson.com/project/e7149d46-fad6-477b-a91c-77e76723df7d) ## Experience - **InsideTracker** — VP, Product & Design (2026—). Leading product strategy and design for a preventative-health platform, translating biomarkers and scientific insight into useful guidance. - **KRING** — Head of Product & UX (2025–26). Product vision, UX strategy, and design across pre-seed ventures in health, sustainability, and energy. - **Included Health** — Staff Design Technologist, AI (2024–25). Built and evaluated voice, real-time video, and multimodal AI prototypes. Coached product and design teams in AI-enabled workflows. - **Noodle AI** — Co-founder (2023–25). Raised angel funding, assembled a team, and designed and co-built a personal medical-history product. The company has since dissolved. - **Obvious Ventures** — Designer in Residence (2018–19). Advised founders on product strategy, research, design hiring, and team development. Led UX for DarwinAI’s explainability product. - **Grand Rounds** — Lead → Principal → Design Director (2014–18). Built and led the Product Design and Research function across physician tools, consumer health experiences, and internal care platforms. - **Gigwalk** — Head of Design (2012–14). Directed product design, brand, research, and design systems for a distributed mobile-workforce platform. - **Salesforce** — Senior Experience Designer, Mobile (2011–12). Led a mobile application redesign and explored multimodal voice interaction with an advanced-concepts team. - **Google** — Mobile User Interface Designer (2007–10). Designed early touch interfaces across Search, Local, Calendar, and other products. Led ethnographic research in Beijing and Chongqing. - **Yahoo!** — Interaction Designer, Mobile (2005–07). Designed the Mobile Portal and experiences spanning Messenger, Local, oneSearch, and Photos. - **Independent & studio work** — Design, technology & interactive media (Earlier). Interactive work for Smithsonian, Lincoln Center, Dr. Seuss, IBM, Bose, Caterpillar, PepsiCo, John Hancock, Lockheed Martin, SoftBank Japan, and Harvard Medical School. Alongside these roles: independent product design, engineering, and advisory work since 2019. Experience drawn from Will’s executive résumé. ## Selected work - [Communicable Disease Dashboard (Public Health)](https://dzierson.com/project/a29df936-41db-4244-b968-203b71235c76) — The Public Health Company, 2025. **See the risk. Know where to act.** Turning complex disease signals into a clear operational picture for global enterprises. - [Google Beijing: Project Backpack (校园 背包客)](https://dzierson.com/project/dff59819-cbc1-46ab-8cef-011e0e2a963f) — Google, 2007. **A microethnography with strategic imperative.** Three months of field research, team mentorship, and product exploration with Google Beijing. - [Human Knowledge Preservation Initiative](https://dzierson.com/project/04193cf2-c3f0-4698-89d5-7593e1f29563) — Project Ariadne (The Ariadne Foundation), 2025. **Inspiration from the distant past to inform the distant future.** An exploration of textiles, materials science, and the preservation of human knowledge. - [AI Healthcare Companion](https://dzierson.com/project/e7149d46-fad6-477b-a91c-77e76723df7d) — Noodle AI (Co-founder), 2023. **An AI companion for complex care.** Co-founded and co-built a platform that brought scattered health records into a patient’s own conversational workspace. - [Smithsonian Guide Device (SiGuide)](https://dzierson.com/project/7b4a9961-a088-46f4-bc88-8f0d64589227) — The Smithsonian Institution, 2005. **A location-aware guide for the Smithsonian.** A location-aware guide, designed and developed for Smithsonian visitors. - [Agentic OS](https://dzierson.com/project/d8307ba2-cc0e-4fc5-bc00-48fd5458dd08) — Confidential Client, 2025. **A browser-based OS built with AI. From the ground up.** A working exploration of persistent agents, shared memory, and natural-language interaction. ## All work (22 case studies) - [Gisted](https://dzierson.com/project/b1a2c3d4-e5f6-7890-abcd-111111111111) — Vibrant Labworks (Founder), 2026. Don't just send a link. Send the gist. Gisted turns links, videos, documents, and ideas into pages people actually open and respond to. - [Smart, in the moment saving](https://dzierson.com/project/f0163fc8-9e6a-4970-a2e0-b420a208402f) — Stash (Founder & Developer), 2026. Stash (gostash.it) is a personal digital catchall that automatically enriches URLs, PDFs, images, Reels/Tiktoks, screenshots, videos, voice memos, markdown and more. - [Agentic OS](https://dzierson.com/project/d8307ba2-cc0e-4fc5-bc00-48fd5458dd08) — Confidential Client, 2025. AgenticOS was an early experiment in AI-native productivity — a browser-based OS-like environment where professionals could build and run agentic, always-on applications without requiring full control of a physical… - [AI Healthcare Companion](https://dzierson.com/project/e7149d46-fad6-477b-a91c-77e76723df7d) — Noodle AI (Co-founder), 2023. Founded in 2023, Noodle started as a RAG-based note-taking app and evolved into something more ambitious: a platform designed to give patients real ownership of their health information. - [Smithsonian Guide Device (SiGuide)](https://dzierson.com/project/7b4a9961-a088-46f4-bc88-8f0d64589227) — The Smithsonian Institution, 2005. SiGuide is a location-aware, handheld museum guide created for several of the museums at the Smithsonian. It combines audio, video, and rich media to support self-guided exploration of the collection. - [Human Knowledge Preservation Initiative](https://dzierson.com/project/04193cf2-c3f0-4698-89d5-7593e1f29563) — Project Ariadne (The Ariadne Foundation), 2025. Project Ariadne merges ancient Indigenous weaving techniques with advanced materials science to create an innovative system for long-term knowledge preservation. - [Multimodal AI Transformation](https://dzierson.com/project/cc5b34e1-e0a7-4a59-b3b6-c9e814585c34) — Included Health, 2025. At Included Health, I designed and prototyped a number of AI-based experiences to explore how natural language interfaces could streamline patient intake, streamline physician workflow, reduce documentation time, and… - [Dash AI Browser (MacOS app)](https://dzierson.com/project/1c0304f9-0d6c-4c4d-a190-47a32b18413c) — Self-started, 2025. Started in 2025 as a self-directed research project, Dash began with a straightforward question: what would a browser actually look like if AI were native to it — not bolted on as a sidebar, but woven into how the page… - [GenSynth ML Explainability Tool](https://dzierson.com/project/af307088-2085-4a60-b697-964b6b37eb82) — Darwin AI, 2019. DarwinAI’s GenSynth platform was built to address the deep learning “black box” problem by making convolutional neural networks more transparent and more efficient. - [Expert Physician Care Portal](https://dzierson.com/project/d409e162-fadd-48c6-b669-19bd8c610a35) — Grand Rounds (Now Included Health), 2025. Grand Rounds (now Included Health) was built on a compelling but operationally demanding premise: connect patients facing serious diagnoses with the country's top medical specialists for expert second opinions. - [AI Clinical Care Knowledge Tool](https://dzierson.com/project/de35cf67-0aec-4984-b320-4883d8e5a0d0) — Included Health, 2025. Included Health has the unique challenge of having millions of members to cater to across a number of offerings including direct line of site care (primary care), longitudinal health, healthcare navigation, billing… - [Google Beijing: Project Backpack (校园 背包客)](https://dzierson.com/project/dff59819-cbc1-46ab-8cef-011e0e2a963f) — Google, 2007. In 2010, I was asked to head to Beijing for an extended, three-month project with two distinct aims: supplement and mentor the relatively junior Google Beijing design team, and propose a product initiative that might… - [Google Mobile Search & Apps](https://dzierson.com/project/0265fc32-d8b6-4951-96f6-789fb9b107cc) — Google, 2010. At Google, I led design on the Mobile Search team during one of the most significant platform shifts in the company's history — the move from desktop to touch-based mobile interfaces. - [Legendary Hollywood Photographer Portfolio](https://dzierson.com/project/fe372618-4767-448d-95b7-5089e3ec86c0) — B/H, 2022. Dr. Bruno Bernard, or “Bernard of Hollywood,” was the most sought-after photographer of Hollywood’s Golden Era. He has been coined the “King of Glamour” and the Rembrandt of Photography. - [Social Media Profile Aggregator](https://dzierson.com/project/f7779201-ce08-49d3-9786-a4a5090f0828) — ThisIsMe, 2010. This is me was a social media profile aggregator, meant to allow high-profile individuals with multiple social media presences across the web to aggregate them in one place (and a single URL). - [Visual Storytelling Assistant](https://dzierson.com/project/ec78bbba-eb18-48b4-b912-255263a007eb) — Storyblocks AI, 2023. The potential for AI to assist with the “blank page problem” writers often encounter is an interesting proposition. - [Open Sourced Browser-based OS](https://dzierson.com/project/7d67278a-09bd-4ab2-bc3d-794e97e43d69) — Self-Started, 2026. Trying building more in public this year. First up: a React library for building macOS-style windowed desktop environments in the browser. - [Communicable Disease Dashboard (Public Health)](https://dzierson.com/project/a29df936-41db-4244-b968-203b71235c76) — The Public Health Company, 2025. The Public Health Company was built around a clear and urgent problem: global enterprises — Fortune 50 companies, hospitals, hotel chains, manufacturers — have almost no visibility into communicable disease risk at the… - [Optum Store](https://dzierson.com/project/ab7e77cd-080f-43c8-b583-2d69a83f1509) — Optum, 2022. At Optum, I partnered with the design and product teams to refine the homepage experience, ensuring it better aligned with user needs and business goals. - [Stickerbooth (Photoboth App with Stickers)](https://dzierson.com/project/b99632e2-7837-43cc-8d6f-714b05d1d7da) — Self started, 2011. I built an app back in 2011 that allowed people to put stickers on their pictures. It was a fun little thing. - [Gigwalk Consumer and Admin Apps](https://dzierson.com/project/38d70eab-b25c-46ee-b3c7-2c443d421f65) — Gigwalk, 2013. Gigwalk is a mobile app platform that connects businesses with on-demand workers, or "Gigwalkers," to complete various tasks in a specific location. - [Crescendo](https://dzierson.com/project/679a3f8d-d191-45b3-bfa9-3a5f6fd41047) — Crescendo, 2022. Improving medical research and enabling better health outcomes Crescendo was founded by Sam Roosz and Michael Glassman to enable better health outcomes for patients, while protecting the interests of research… ## Writing I’m interested in how people understand, govern, and work with intelligent systems—and in bringing a builder’s perspective to teaching and research. - [Agentic AI Is Not a Magical Elf](https://medium.com/@willdzierson/agentic-ai-is-not-a-magical-elf-ad1ae273493c) — Perspective, Medium - [AI Agent Accountability: Who’s Responsible When No One’s Home?](https://medium.com/@willdzierson/ai-agent-accountability-whos-responsible-when-no-one-s-home-2298323117d5) — Essay, Medium ([read it here](https://dzierson.com/essay/31ce52f5-d961-42e7-9b6b-e8eaafceee41)) - [Building a RAG-Based Portfolio Chatbot](https://medium.com/@willdzierson/building-a-rag-based-portfolio-chatbot-with-lovable-dev-supabase-and-the-openai-api-0ceafc7a0d9a) — Build notes, Medium - [On OpenAI's new open-source models (gpt-oss) & Healthcare Implications](https://dzierson.com/essay/cba7743e-421e-4dd4-adb9-787ff415fd80) — Essay, August 7, 2025 - [AI/LLM on-device capability & information privacy](https://dzierson.com/essay/7923ee58-0e7b-4a55-9ec1-5ccc37bf7417) — Essay, June 10, 2025 - [On communicating information through deep time](https://dzierson.com/essay/1b5b308a-a372-4f55-80d9-7b7b1c2bee58) — Essay, June 5, 2025 - [Book Review: AI Fairness](https://dzierson.com/essay/7481245d-30a5-479c-9431-78e82d032c5f) — Essay, June 2, 2025 - [AI Governance for Product & Design folks](https://dzierson.com/essay/c843d15e-5d49-4283-b8b8-4a4d0f0a9ef3) — Essay, May 30, 2025 - [The Urgency of Interpretability](https://dzierson.com/essay/500d1e86-b3a6-4fdc-954b-1f5835814c30) — Essay, May 15, 2025 - [Reading When Human-AI Interactions Become Parasocial ](https://dzierson.com/essay/83669b98-eb9c-4a34-81fc-bf90e1edc920) — Essay, May 9, 2025 - [My thoughts on the future of AI](https://dzierson.com/essay/4b02f3dd-357d-4711-aa4b-a4e30e300937) — Essay, May 6, 2025 Published contributor, Mobile Persuasion — Stanford Persuasive Technology Lab. Portfolio [featured by Lovable](https://lovable.dev/guides/11-ux-portfolio-examples). ## Selected teams and clients Google, Included Health, SoftBank, Salesforce, Yahoo!, Smithsonian, Optum, IBM, Bose, Caterpillar, Obvious Ventures, Lincoln Center, Lockheed Martin, Dr. Seuss, InsideTracker. ## Contact [will@dzierson.com](mailto:will@dzierson.com) · Saratoga Springs, New York · [LinkedIn](https://www.linkedin.com/in/will-dzierson-1081963/) · [Medium](https://medium.com/@willdzierson) --- # Career — Will Dzierson **A builder’s perspective. At every scale.** From early mobile products to AI-native ventures. A career moving between making the product and helping teams do their best work. - **InsideTracker** — VP, Product & Design (2026—). Leading product strategy and design for a preventative-health platform, translating biomarkers and scientific insight into useful guidance. - **KRING** — Head of Product & UX (2025–26). Product vision, UX strategy, and design across pre-seed ventures in health, sustainability, and energy. - **Included Health** — Staff Design Technologist, AI (2024–25). Built and evaluated voice, real-time video, and multimodal AI prototypes. Coached product and design teams in AI-enabled workflows. - **Noodle AI** — Co-founder (2023–25). Raised angel funding, assembled a team, and designed and co-built a personal medical-history product. The company has since dissolved. - **Obvious Ventures** — Designer in Residence (2018–19). Advised founders on product strategy, research, design hiring, and team development. Led UX for DarwinAI’s explainability product. - **Grand Rounds** — Lead → Principal → Design Director (2014–18). Built and led the Product Design and Research function across physician tools, consumer health experiences, and internal care platforms. - **Gigwalk** — Head of Design (2012–14). Directed product design, brand, research, and design systems for a distributed mobile-workforce platform. - **Salesforce** — Senior Experience Designer, Mobile (2011–12). Led a mobile application redesign and explored multimodal voice interaction with an advanced-concepts team. - **Google** — Mobile User Interface Designer (2007–10). Designed early touch interfaces across Search, Local, Calendar, and other products. Led ethnographic research in Beijing and Chongqing. - **Yahoo!** — Interaction Designer, Mobile (2005–07). Designed the Mobile Portal and experiences spanning Messenger, Local, oneSearch, and Photos. - **Independent & studio work** — Design, technology & interactive media (Earlier). Interactive work for Smithsonian, Lincoln Center, Dr. Seuss, IBM, Bose, Caterpillar, PepsiCo, John Hancock, Lockheed Martin, SoftBank Japan, and Harvard Medical School. Alongside these roles: independent product design, engineering, and advisory work since 2019. Experience drawn from Will’s executive résumé. ## Selected teams and clients Google, Included Health, SoftBank, Salesforce, Yahoo!, Smithsonian, Optum, IBM, Bose, Caterpillar, Obvious Ventures, Lincoln Center, Lockheed Martin, Dr. Seuss, InsideTracker. --- # Gisted Vibrant Labworks (Founder), 2026 · Case study by Will Dzierson **Tags:** AI, SaaS, React, Supabase **Project link:** [https://gisted.co](https://gisted.co) ## My contribution **Sole designer and engineer.** Conceived the product, designed every screen, built the full stack — React frontend, 40+ Supabase edge functions, enrichment pipeline, Stripe billing, Chrome extension, video generation. From first sketch to production SaaS. ## The work Don't just send a link. Send the gist. Gisted turns links, videos, documents, and ideas into pages people actually open and respond to. Paste any content — a YouTube video, a book, a slide deck, a PDF, a TikTok, a recipe — and Gisted enriches it with AI: summaries, key moments, full transcript, mentions, diagrams, and suggestions for what else to add. The person's perspective comes first. Your take is the first thing readers see — not a robot summary, but your words, your context, your reason for sharing. A three-step creation flow — Get the gist, Customize, Send — takes a bare URL to a rich, shareable page in minutes. Each content type gets its own treatment: video workspaces with key moments, slide presentations, book summaries, health article breakdowns. Feedback is built in. Reactions, comments, ratings, polls, date pickers, grocery lists. Share via link, email, LinkedIn, or X, and see who opened it and what they thought. The system runs on 40+ Supabase edge functions with a server-side enrichment pipeline, Stripe billing, Google and Apple sign-in, a Chrome extension, and video generation from enriched content. Designed, built, and operated end to end. ## Images and video - [Gisted homepage](https://dzierson.com/portfolio/gisted/home.png) - [AI enrichment board — summaries, key moments, transcript, mentions, diagrams](https://dzierson.com/portfolio/gisted/enrich.png) - [Page editor with suggestion rail and Ask Gisted AI](https://dzierson.com/portfolio/gisted/customize.png) - [Sharing step — social post preview, feedback toggles](https://dzierson.com/portfolio/gisted/send.png) - [Dashboard with rich content type cards](https://dzierson.com/portfolio/gisted/dashboard.png) - [Video workspace](https://dzierson.com/portfolio/gisted/video.png) - [Reader view with conversation](https://dzierson.com/portfolio/gisted/reader.png) - [Slide deck presentation](https://dzierson.com/portfolio/gisted/slides.png) --- # Smart, in the moment saving Stash (Founder & Developer), 2026 · Case study by Will Dzierson **Tags:** LLMs, Machine Learning, AI, Software Development **Project link:** [https://gostash.it/](https://gostash.it/) ## My contribution As founder, developer and designer, I owned the full lifecycle of Stash: defining the product vision, mapping use cases, and shaping the UX across web and mobile. I designed the card and capture model, implemented the backend, wired up AI enrichment and conversational search, and built the apps. ## The work Stash (gostash.it) is a personal digital catchall that automatically enriches URLs, PDFs, images, Reels/Tiktoks, screenshots, videos, voice memos, markdown and more. Stash employs a series of agents in the background which quickly capture & summarize content related to an item, and stores it in a vector table. Everything becomes semantically searchable and ready for downstream AI workflows. Users can share items with Stash via a Chrome plugin (single button sharing), mobile app (share from any app that supports the share sheet), and MCP. When they need to find something, they chat with their own information layer to summarize, cross-reference, and rediscover content—whether that’s reading notes, health logs, travel ideas, or restaurant recommendations. This makes Stash function as a low-friction, always-available capture surface that turns daily inputs into durable, portable knowledge. You never know when that thing you saw and took a screenshot of will come in handy again. I built Stash end-to-end in collaboration with Claude & Codex, covering product design, agentic engineering, backend, and the various apps (Chrome, iOS, MacOS). The stack runs on Supabase & Vercel, with fly.io hosting a Hermes implementation for durable tasks & continuous improvement, as well as enrichment agents. ## Images and video - [Stash is an information enrichment platform that works across most media types: links, videos, docs, markdown, audio files, etc. Agents are deployed in the background to gather as much additional information about the object as possible, and this information is made searchable and practically useful to the user.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/blgs5namhm.png) - [The Stash UI -- live. Sign up at https://gostash.it/](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/cj7b14qrjx.mov) - [The various ways to Stash: Chrome extension, mobile app, and MCP.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/5wb1cf5dfj.mov) - [A demonstration of Stash's enrichment capabilities from the Stash homepage. Paste a link in: it works!](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/v8wszfszx5.mov) - [A demonstration of Stashing various items using the share button from most mobile apps. Stashing can be done from anywhere.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/o0myz56if9.mov) - [A demonstration of Stash's MCP capabilities.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/5b19p7hlg4l.mov) --- # Agentic OS Confidential Client, 2025 · Case study by Will Dzierson **Tags:** UX, Healthcare, LLMs, Machine Learning, Speculative Future, AI **Project link:** [https://os.dzierson.com/](https://os.dzierson.com/) ## My contribution Working alongside a confidential client, I conceived, designed, and built AgenticOS from the ground up — responsible for the full stack, from UX and interaction design to implementation in TypeScript and Next.js. I integrated Supabase and Postgres/PGVector for the persistent memory and RAG layer, OpenAI for NLU and agentic task execution, Hugging Face for model inference, and built the entire system using Cursor and Claude as core development tools. ## The work AgenticOS was an early experiment in AI-native productivity — a browser-based OS-like environment where professionals could build and run agentic, always-on applications without requiring full control of a physical machine. Built about a year before OpenClaw debuted, it served as a proving ground for what an AI-first OS could look like. AI wasn't bolted on as isolated features — it was the foundation. Every layer of the experience was designed around intelligent automation, natural language interaction, and persistent agentic task execution. Core explorations included a cross-app memory layer that made relevant knowledge instantly available to any application without requiring complex integrations, and a visual layer giving applications real-time awareness of on-screen content for true OS-wide multimodal capability. Core Applications Productivity Suite: - **Notes** - Rich text editor with AI-powered RAG chat for intelligent note searching and content generation - **Text Editor** - Full-featured document editing with syntax highlighting - **Calculator** - Standard mathematical operations - **Mail** - Email management with Resend integration - **File Manager** - Complete file system navigation and management Creative Tools: - **Paint** - Canvas-based drawing application with Fabric.js - **Photo Editor** - Advanced image editing with AI assistant for automated enhancements - **Photo Viewer** - Image gallery and viewing capabilities - **Music Player** - Audio playback with metadata support Development & System: - **Terminal** - Command-line interface for system operations - **App Builder** - AI-powered application creation wizard - **Agent Creator** - Build custom AI agents and workflows - **System Preferences** - OS configuration and settings - **Documentation** - Integrated help and API documentation AI-Enhanced Apps: - **Browser** - Web browsing with AI chat panel and agentic task execution (e.g., "find cheapest flights to NYC") - **OSChat** - Universal AI assistant with voice command integration + agentic abilities (orchestration, service discovery) - **Oracle** - Advanced AI consultation and analysis - **Willbot Game** - Interactive AI character gaming experience AI Features & Capabilities Intelligent Command Layer: - **Voice command** processing and natural language understanding - **Agentic task** execution ("open browser and find me cheap flights") - **Cross-app** intelligent routing and coordination AI-Powered Content: - **Notes** with RAG-based AI chat for content discovery - **Photo editing** with AI enhancement suggestions - **Browser** with specialized agents (flight search, general web tasks) - **Real-time** AI assistance across applications via realtime vision Modular AI Architecture: - **Custom** agent creation and deployment - **App-specific** AI command registration - **Intelligent** workflow orchestration - **Background** AI processing and sync Voice & Natural Language: - **Continuous voice** chat capabilities - **Voice command** execution across all apps - **Natural language** to system action translation ## Images and video - [The OS login screen, featuring a custom Spline-rendered background — the first signal that this wasn't a conventional productivity tool.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/m6txi8da39.png) - [\[VIDEO\] A walkthrough of several AgenticOS capabilities: the Notes app with natural language search powered by the OS memory layer and RAG, cross-app interaction demonstrating the OS' ability to understand intent, the in-OS browser for cross-app interactions, and the photo editor (AgenticOS Photoshop).](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/aryznoepnp.mp4) - [\[VIDEO\] A technical view of the OS with debugging tools enabled (Chrome console and AgenticOS' own debugging panel). The LLM app is communicating across apps and using the vision layer of the OS to assess the current activity and take appropriate action. To protect patient data, the LLM refuses to perform an analysis. In a production environment, this would invoke a specialized model to provide radiology results.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/e6jgzmuwmh.mp4) - [The AgenticOS desktop — intentionally minimal, keeping focus on the applications and the agentic layer running beneath them.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/7zbdqevh712.png) - [The prompt-based App Builder I designed and built into AgenticOS. A user describes what they need, and the OS generates and compiles a working app in a sandboxed environment.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ox13xdkym7.png) - [Apps could be configured and calibrated upfront, then generated with a single click — collapsing the gap between idea and working software.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/couw1mdggar.png) - [An app produced with the app builder for a materials sciences use case, using Reactflow to enable infinite canvas interactions — demonstrating that the platform could accommodate complex, domain-specific tooling.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/hagmyjbuqb.png) - [The TypeScript output generated by the App Builder. Each app compiles down to a handful of files loaded dynamically via Next.js — making AI-generated apps first-class citizens of the OS.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/6mu4b2q0ql.png) - [The in-OS browser, built with RAG and page-scraping to let users ask questions about whatever's in their viewport — summarize a page, extract a shopping list, or trigger agentic tasks. This became the foundation for the MacOS-based Dash browser I built afterward.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/1k1b8uxtpqv.png) - [The Notes app showing multitasking in context — notes are stored and retrievable conversationally via RAG, letting users find and surface content through natural language rather than manual search.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ooybu8aywu.png) - [The OS visual layer in action: the agentic layer can interpret what's currently visible on the desktop and take coordinated action, enabling a shared context between the user and the OS.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/zal79inynss.png) --- # AI Healthcare Companion Noodle AI (Co-founder), 2023 · Case study by Will Dzierson **Tags:** TypeScript, UX, Prototyping, LLMs, AI, Software Development **Project link:** [https://www.noodleai.app/videos/noodle\_walkthrough\_2.mp4](https://www.noodleai.app/videos/noodle_walkthrough_2.mp4) ## My contribution I co-founded Noodle and led the end-to-end design and co-lead engineering. I designed the app and co-built the platform using Next.js (TypeScript), Tailwind, and Supabase/Postgres with PGVector for RAG, along wtih contract engineers. For AI, we integrated multiple LLM engines — OpenAI, Hugging Face (Meditron-7B/Mistral), and Claude — depending on the task. Added Twilio for WhatsApp-based capture, Deepgram for voice transcription, and RapidAPI for image tagging, OCR, and recognition. The multimodal voice experience was built in Swift. ## The work Founded in 2023, Noodle started as a RAG-based note-taking app and evolved into something more ambitious: a platform designed to give patients real ownership of their health information. The focus shifted to people managing complex, multi-specialist conditions — the kind that generate scattered records, forgotten instructions, and fragmented data across providers. The core insight was that health record balkanization is the root problem. Noodle addresses it by letting people capture and consolidate health events through whatever channel is easiest — web app, WhatsApp, voice, or images — and then surfaces that information through a conversational UI, making it genuinely useful rather than just stored. Working with my co-founder, Julz Bautista, we raised an angel round and generated significant interest from partners and members of the healthcare community. The platform leverages LLMs, vector search (RAG), personal health journaling, health record consolidation, and self-sampling to help surface actionable insights and enable more meaningful interactions between patients and their physicians. Noodle has since dissolved. This is historical work. ## Images and video - [The Noodle marketing site and mobile experience — designed to communicate the core value proposition to patients managing complex, ongoing health conditions.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/tixziglvmcr.png) - [\[VIDEO\] A full walkthrough of the Noodle app — demonstrating the end-to-end experience from logging a health event to asking the assistant a follow-up question and getting a context-aware response. Built in Cursor + Swift for iPhone.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/v9iacssomo.mp4) - [\[VIDEO\] An example use case, where the user is able to easily summon the piece of information they're interested in by asking Noodle to review their after visit summary and share the doctor's name.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/bv92uha00lv.mov) - [The RAG-powered health assistant at work — generating a structured summary of a patient’s pain history and pulling relevant context from their logged records to inform the response.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/e4o2cy2ezzv.png) - [From the investor pitch deck — introducing Noodle’s mission to empower patients to centralize, better interpret, and derive value from their health information.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/31g46hj3dal.png) - [A research slide from our pitch deck framing the core problem — patients routinely deal with too much information that’s rarely centralized, often poorly understood, and too easily forgotten.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/gs6mzc4z83m.png) - [The reality Noodle was built to solve: patient health information is paper-based or scattered across multiple apps, making it nearly impossible for patients to recall, share, or act on their own data](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/0tmlpbxy1o3.png) - [Market research backing the problem space — 83% of patients forget critical health information after appointments, with $236B in costs tied to patients unable to process their own information and $100B lost annually to non-adherence.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/0qmy6hxvp8ir.png) - [A technical architecture diagram showing how Noodle uses NER tokenization to extract and tag structured entities from unstructured patient input — enabling more precise RAG retrieval and clinical context-building.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/qrmi5uwv72g.png) - [User discovery work mapping a pregnancy persona — one of several complex condition scenarios we used to define what "managing your own health knowledge" actually looks like in practice.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/i6h24vd580g.png) --- # Smithsonian Guide Device (SiGuide) The Smithsonian Institution, 2005 · Case study by Will Dzierson **Tags:** Mobile Design, Prototyping, UX ## My contribution I served as the lead designer and developer for the SiGuide user interface, creating a Flash-based application optimized for PocketPC devices running Windows CE. My responsibilities spanned interaction design, visual design, and front-end implementation, including navigation models, tour and artifact views, and the interactive 3D tour-path visualizations. I also collaborated closely with museum stakeholders to translate curatorial goals into clear, accessible UI patterns suitable for first-time technology users. ## The work SiGuide is a location-aware, handheld museum guide created for several of the museums at the Smithsonian. It combines audio, video, and rich media to support self-guided exploration of the collection. It runs on a PocketPC-style device and was designed to feel intuitive for first-time technology users while still offering depth and control for more experienced visitors. The interface centers on a clear home screen that exposes the main modes of the experience—interactive map, gallery guide, today’s events, tours, and activities—so visitors can quickly choose how they want to navigate the museum. Individual tours, such as “Letters and Letter Writing,” present a concise synopsis, a visual overview of key objects, and an isometric tour path that orients visitors in physical space. Each tour stop is anchored by a specific artifact (for example, historic maps or a Concord-style stagecoach), paired with contextual audio and “About” or “Explore” panels that let visitors dive deeper into the story of each object. Across the system, SiGuide emphasizes a simple browsing model with forward/back navigation, a persistent home/map bar, and features like “Locate on Map” and “Scrapbook it!” that help visitors save and relocate favorite artifacts. The result is a cohesive, narrative-driven experience that turns the museum visit into an accessible, media-rich tour where visitors can either follow a structured path or wander and let the device recognize where they are and surface relevant content. ## Images and video - [The SiGuide experience was delivered on a custom handheld Windows CE device that visitors checked out from a dedicated kiosk at the entrance to participating Smithsonian museums. The ruggedized unit featured a landscape touch screen, integrated speakers, and physical controls sized for comfortable use while walking the galleries. It was deployed across multiple sites—including the Udvar-Hazy Air and Space Museum, the National Museum of Natural History, and the National Postal Museum—so the same core hardware and UI could support different collections and tour content.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/4jn2ard683w.png) - [Smithsonian Guide Device (SiGuide) — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/zf5cy7p6aqh.png) - [Smithsonian Guide Device (SiGuide) — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/wgovb0jy5f.png) - [Smithsonian Guide Device (SiGuide) — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/65kq18kxfxv.png) - [Smithsonian Guide Device (SiGuide) — image 5](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/f9cj12axpwm.png) - [Smithsonian Guide Device (SiGuide) — image 6](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/sshrkxg326.png) --- # Human Knowledge Preservation Initiative Project Ariadne (The Ariadne Foundation), 2025 · Case study by Will Dzierson **Tags:** Materials Science, UX, LLMs, Speculative Future, AI **Project link:** [https://www.projectariadne.info/](https://www.projectariadne.info/) ## My contribution I originated the project, with supervision and input from Dr. Vint Cerf and others who are experts in the area of knowledge preservation, linguistics, and materials science. ## The work Project Ariadne merges ancient Indigenous weaving techniques with advanced materials science to create an innovative system for long-term knowledge preservation. Through AI-driven material simulations, we're developing ultra-durable synthetic fibers designed to withstand millennia while honoring Indigenous textile wisdom. ## Images and video - [Human Knowledge Preservation Initiative — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/1h3hhvebpty.png) - [Human Knowledge Preservation Initiative — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/89n4et37ale.png) - [Human Knowledge Preservation Initiative — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/6bpiadpdish.png) - [Human Knowledge Preservation Initiative — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/litwwypb62.png) - [Human Knowledge Preservation Initiative — image 5](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/4qii6s8b21f.png) - [Human Knowledge Preservation Initiative — image 6](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/da00xh4cfhh.png) - [Human Knowledge Preservation Initiative — image 7](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/0le0j1ry02y9.png) - [Human Knowledge Preservation Initiative — image 8](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/5fkaim13rbn.png) - [Human Knowledge Preservation Initiative — image 9](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/6ngujez53cn.png) - [Human Knowledge Preservation Initiative — image 10](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ofse4hw4hq.png) --- # Multimodal AI Transformation Included Health, 2025 · Case study by Will Dzierson **Tags:** UX, Prototyping, LLMs, Machine Learning, AI **Project link:** [https://y4qwrjnnsy2vya5v.vercel.app/](https://y4qwrjnnsy2vya5v.vercel.app/) ## My contribution As the lead on this initiative, I was responsible for the full prototype lifecycle — from conversation flow design and prompt engineering to TypeScript/Vercel implementation with OpenAI and ElevenLabs. I collaborated directly with clinical, product, and leadership stakeholders to validate and ship the experience, and championed AI-assisted design workflows across the team, introducing tools like v0, Cursor, Lovable, and GitHub Copilot. ## The work At Included Health, I designed and prototyped a number of AI-based experiences to explore how natural language interfaces could streamline patient intake, streamline physician workflow, reduce documentation time, and improve accessibility. The goal was to evaluate whether a conversational experience could make healthcare interactions more human, intuitive, and efficient — especially in high-friction, high-stakes environments like telehealth and care navigation. I led the end-to-end experience design, from conversation flow mapping and prompt engineering to interface prototyping and voice synthesis integration. The prototype supported real-time, conversational data capture and retrieval, enabling users to speak naturally and receive intelligent, context-aware responses from the assistant. To bring the prototype to life, I used Cursor and v0 to build the interface, and integrated OpenAI's language models for NLU and ElevenLabs for synthetic voice output, creating a lifelike interaction loop. I collaborated with clinical ops and engineering stakeholders to align the voice experience with real-world care workflows and regulatory considerations. The project was always dual-purpose: redesign a broken flow, and learn something meaningful about patient preferences for AI in healthcare. We ran a structured Maze experiment to test all three experiences (original UI, chat-based LLM, and voice-based LLM) head-to-head, and voice came out ahead by a wider margin than anyone anticipated — a finding that shifted how the team thought about conversational AI across the board. The research directly informed the design language and interaction patterns for AI experiences across several subsequent projects at Included Health, establishing a foundation for how the team approached conversational UI going forward. ## Images and video - [\[VIDEO\] A refined version of the chat-based and voice-based intake experiences, built with v0 (Vercel), OpenAI for NLU, and ElevenLabs for TTS and voice synthesis.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/sa164xief28.mp4) - [\[VIDEO\] An early prototype of the voice-based intake experience, built with Spline and ElevenLabs' API. The UI responds in real time to user input — this demo helped generate stakeholder buy-in for deeper voice exploration.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/9yrf64krltu.mp4) - [The conversational intake flow in action, showing how the assistant guides patients through onboarding via natural dialogue.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/r4vmefuwwyq.png) - [I moved quickly from Figma flows into code using v0, Lovable, and Cursor — enabling higher-fidelity interactions and hands-on experimentation with voice and avatar-based interfaces.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/bdnrpq9l1u.png) - [I reimagined the e-opinion onboarding flow — a multi-screen experience suffering close to 100% attrition — by designing three new intake approaches: an AI-enhanced flow, a chatbot-based experience, and a voice-based experience.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/fdeoh6isnp.png) --- # Dash AI Browser (MacOS app) Self-started, 2025 · Case study by Will Dzierson **Tags:** LLMs, AI, Software Development **Project link:** [https://drive.google.com/file/d/1MZq9JfzpkqhizayD2lH--eWDjMxQpfIV/view?usp=sharing](https://drive.google.com/file/d/1MZq9JfzpkqhizayD2lH--eWDjMxQpfIV/view?usp=sharing) ## My contribution I designed and built Dash end-to-end in Xcode using Swift and WebKit. The AI layer is built on local embeddings for semantic page understanding, with Claude and OpenAI handling inference depending on the task. I used Cursor as a primary development tool throughout. For data and agent orchestration, I integrated Supabase (Postgres + pgvector) for vector storage and auth, MCP-based agents for tool invocation, and Perplexity for supplementary browsing context. The result is a fully working native macOS prototype built entirely solo. ## The work Started in 2025 as a self-directed research project, Dash began with a straightforward question: what would a browser actually look like if AI were native to it — not bolted on as a sidebar, but woven into how the page is understood and acted upon? Inspired by the emerging wave of AI-first browsers like Dia, I set out to design and build a working answer. The result is a Safari-based macOS app — internally nicknamed "galliano" — that reads the page you're viewing, builds a semantic model of its content, and uses that understanding to take action on your behalf. It can summarize any page or site, answer contextual questions, and spin up specialized agents to handle tasks like finding better prices for products, generating grocery lists from recipes, checking business ratings, or placing a DoorDash order. It also supports "meta" actions like emailing you a tailored summary of the current page or extracting and organizing specific information — making it a genuinely useful companion for everyday research and shopping workflows. Under the hood, Dash uses a locally available model to embed the current page and selected links, giving the AI a semantic foundation to decide which agents or tools to invoke. It combines local and remote retrieval-augmented generation with Supabase for vector storage and auth, MCP-based agents, and a suite of APIs to orchestrate tasks like pricing lookups and local retail discovery. I built the browser end-to-end in Xcode using Swift and WebKit, with Claude, Cursor, and OpenAI as the core development and inference tools. The outcome is a working prototype that demonstrates how agentic workflows, local embeddings, and modern LLM tooling can be composed into a coherent, native macOS experience — and a proof of concept for what AI-native browsing might actually feel like. The repository and a demo are available for anyone interested in running or extending the project. The codebase includes setup instructions for Supabase, OpenAI, and the required local tooling. ## Images and video - [A side-by-side view of Dash's agentic pipeline — Swift code on the left showing the entity extraction and summarization functions, alongside a live email interaction on the right where Dash surfaces a tailored summary of the current page.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/a4cnoyqc8u6.png) - [\[VIDEO\] A full walkthrough of Dash in action — demonstrating how the browser reads and understands the current page, invokes specialized agents, and completes tasks like price lookups, grocery list generation, and placing a DoorDash order, all from within a native macOS interface.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/g8q4few8xs.mp4) - [Dash navigating the Love & Lemons recipe site — a real-world example of the grocery list agent at work, where Dash reads the page content, identifies ingredients, and compiles a structured shopping list without any manual input from the user.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/wd9rviexfkp.png) - [A closer look at the development workflow — Swift agent code on the left alongside an active Cursor session on the right, showing the iterative process of building and validating Dash's agentic capabilities with Claude as a coding collaborator.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/s283jylp1t.png) --- # GenSynth ML Explainability Tool Darwin AI, 2019 · Case study by Will Dzierson **Tags:** UX, Prototyping, Machine Learning, AI ## My contribution Principal Designer. I led UX research, interaction design, prototyping, and visual design for the project, mapping the ML workflow and shaping the core explainability and performance-analysis experience in close collaboration with the DarwinAI team. ## The work DarwinAI’s GenSynth platform was built to address the deep learning “black box” problem by making convolutional neural networks more transparent and more efficient. For this project, I designed an interface that helped ML practitioners define confusion matrices, inspect misclassified samples, and trace how CNNs learned from training images, with a particular focus on self-driving vehicle perception workflows. The experience surfaced neuron- and layer-level signals that made it easier to understand which features were driving specific decisions, identify bias and edge cases, and refine models through iteration rather than treating them as opaque systems. The result was a more usable environment for exploring experiments, comparing model variants, and moving toward compact, high-performing networks suitable for real-time edge deployment in safety-critical contexts. ## Images and video - [GenSynth overview interface, showing the primary workspace for exploring model behavior and explainability outputs.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/bnlz6d28t69.png) - [Confusion-matrix and misclassification analysis view, designed to help practitioners isolate failure patterns and edge cases.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/lbim39nc4mc.png) - [Layer- and neuron-level inspection view, revealing the internal signals contributing to model predictions.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/l6v6rc01bc.png) - [Experiment comparison workflow, supporting review of model variants and tradeoffs between accuracy, interpretability, and deployment efficiency.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ouma88gy8jo.png) - [Model analysis detail view, helping teams connect training data, activations, and prediction outcomes in a single workflow.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/0mdtik5ezp9k.png) --- # Expert Physician Care Portal Grand Rounds (Now Included Health), 2025 · Case study by Will Dzierson **Tags:** UX, Prototyping **Project link:** [https://www.includedhealth.com](https://www.includedhealth.com) ## My contribution I served as Lead Designer, owning the project from initial research through final delivery. That meant running the user research myself — conducting field observations and in-depth interviews with practicing physicians, internal operations staff, and clinical reviewers to build a grounded understanding of how specialists actually approach a second-opinion case. From there, I developed the IA and interaction model for the redesigned interface, working through multiple rounds of rapid prototyping to validate patterns with the clinical users before anything was handed to engineering. Key design decisions included modular record views that let experts control their own review sequence, progressive disclosure to manage the density of patient history data, and inline evidence tagging that allowed physicians to annotate and reference specific findings directly within their report workflow. I worked in close collaboration with data scientists and engineers throughout the build — staying embedded in the development process to ensure the implementation held fidelity to the design intent. The clinical precision of the output depended on it. ## The work Grand Rounds (now Included Health) was built on a compelling but operationally demanding premise: connect patients facing serious diagnoses with the country's top medical specialists for expert second opinions. The bottleneck wasn't finding the experts — it was giving them a tool worthy of their time and clinical rigor. The internal Expert Interface had grown organically and wasn't keeping pace with the complexity of the cases coming through. I led the end-to-end redesign of the Expert Physician UI — the primary workspace where specialists reviewed patient records, formed clinical opinions, and authored second-opinion reports. The platform served physicians at the top of their fields: orthopedic surgeons, oncologists, cardiologists, and neurologists reviewing cases that often had significant treatment implications for the patient. Getting the UX wrong wasn't just an efficiency problem — it risked degrading the quality of the clinical output itself. The work began with deep user research: field observations, workflow interviews, and heuristic analysis with clinicians and internal operations staff. What emerged was a clear picture of where the interface was fighting expert cognition rather than supporting it — fragmented record navigation, poor information hierarchy, and no tooling to help experts quickly synthesize complex multi-document patient histories. The redesign addressed each of these systematically, introducing modular record views, progressive disclosure patterns, and inline evidence tagging that mapped to how specialists actually think through a case. The redesigned platform measurably improved expert satisfaction and reduced time-to-completion for second-opinion reports, helping Grand Rounds scale its medical insight services without sacrificing clinical depth or quality. ## Images and video - [The redesigned Expert Physician Interface — the primary workspace where top-tier medical specialists review patient records and compose second-opinion reports. The desktop view shows the full patient case with structured medical history, while the tablet view surfaces the integrated medical imaging viewer for radiological review alongside the record.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/4s3piz0kjc9.png) - [A summary slide framing the design challenge, approach, and impact for the Expert Physician Portal redesign. Physicians needed to evaluate complex patient conditions quickly and efficiently but were resistant to interface changes due to familiarity with the existing tool. A human-centered, research-led approach — grounded in field observations and workflow analysis — drove the redesign toward measurably better physician efficiency and patient outcomes.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/txn8qr5eny.png) - [User research output from the field research phase — Expert Physician User Insights. I conducted direct interviews and observations with active specialist physicians to understand their clinical workflow, their mental models when reviewing patient cases, and the specific failure modes of the existing interface. This research directly shaped the information hierarchy and interaction patterns of the redesign.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/n5d51baioe.png) - [Workflow documentation and current-state mapping for the Expert Medical Opinion process — capturing how cases moved through the system before the redesign. This artifact was developed collaboratively with operations staff and physicians during the research phase, and served as the foundation for identifying the biggest gaps between what the workflow demanded and what the tool supported.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/g6xpio6zxut.png) - [A design concept for the Automated Summary Creation Tool — one of the key process innovations that emerged directly from user research. Physicians cited the time required to manually synthesize patient records as a top pain point. This tool was designed, prototyped, and built as a direct response to that insight, automatically generating structured summaries from patient history to let experts focus their cognitive effort where it matters most.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/qkj0tl0mejf.png) - [The end output of the Grand Rounds Expert Physician platform — a polished, patient-facing second opinion report authored by Dr. James Tibone, Professor at the Keck School of Medicine at USC. This is the artifact the entire physician workflow is designed to produce: a clear, credible, and actionable expert opinion delivered to the patient within 10 days. The quality and usability of the Expert Physician UI directly determines the quality of this report.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/p6m4ubpon7s.png) --- # AI Clinical Care Knowledge Tool Included Health, 2025 · Case study by Will Dzierson **Tags:** Healthcare, Prototyping, LLMs, Machine Learning, AI ## My contribution Design Lead, Prototyping, Design Engineer ## The work Included Health has the unique challenge of having millions of members to cater to across a number of offerings including direct line of site care (primary care), longitudinal health, healthcare navigation, billing support, and many others. The volume of calls and documents (which oftentimes includes a patients' full, unedited medical record) can be overwhelming. This is why the KS (Knowledge System) was developed. So that when a patient calls in, the right answers are at the fingertips of the Customer Care Advocate immediately. I designed the system to make full use of LLMs and RAG, simultaneously searching millions of documents — both care-related and patient-specific — to surface the right answer in context. The results were measurable: the system reduced AHT (Average Handle Time) by an average of 1.5 minutes per case, a meaningful cost savings at scale, without compromising the quality of care Included Health is known for. ## Images and video - [The KS home view — Quick Links navigation on the left, with AI-powered AHT reduction suggestions and live Announcements in the right panel.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/y4gagz4b30s.png) - [The KS metrics dashboard — showing real-time answer volume by day, week, and pod, alongside an AHT Improvement Over Time trend chart tracking minutes saved per case.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/u48qy64uigg.png) - [The search interface with an active Recent Questions dropdown — surfacing contextually relevant queries based on the advocate’s current member and insurance context.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/rywzexpfyjq.png) - [The AI-powered answer view — showing a Member Answer and a detailed Care Associate Answer side by side, generated in real time from the member’s plan documents and medical record.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/z3qy1f34qa.png) - [The “Modify Member Answer” modal — allowing advocates to tailor AI-generated responses by adjusting reply language, question complexity, reply length, health literacy level, and tone before sharing with the member.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/yaswi86amv.png) - [A follow-up interaction in the answer view — with the advocate reviewing treatment coverage and provider options, surfaced contextually from the member’s plan and care history.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/qmkbaxis318.png) --- # Google Beijing: Project Backpack (校园 背包客) Google, 2007 · Case study by Will Dzierson **Tags:** UX, Software Development, TypeScript ## My contribution I served as lead designer and co-research lead for the full three-month engagement. Alongside research partner George Zhang, I designed and ran a field research program that included in-person interviews, contextual inquiry sessions, and cultural immersion visits at Chinese university campuses — including a dedicated mini-immersion at Chongqing University to capture the distinct dynamics of Tier 2 city student life. I synthesized field findings into a structured insight framework and led a cross-functional workshop with the Google Beijing team to generate product concepts from the research. I then owned the design and narrative development of the final concept portfolio, producing interaction concepts, scenario illustrations, and a product brief that translated student behavioral insights into actionable design directions for Google's China product teams. ## The work In 2010, I was asked to head to Beijing for an extended, three-month project with two distinct aims: supplement and mentor the relatively junior Google Beijing design team, and propose a product initiative that might meaningfully move the needle on Google's fledgling presence in China. At the time, Google was a distant fourth or fifth in brand recognition against deeply entrenched native incumbents like Baidu — and the challenge of winning Chinese users over wasn't just a marketing problem, it was a product problem. I spent much of the engagement in the field alongside my research partner, George Zhang, conducting immersive research with Chinese university students across Beijing and Chongqing. We spoke at length with students about their hopes, fears, dreams, and desires — particularly through a technological lens. The research took us into dorm rooms, cafeterias, campus common areas, and computer labs, building a textured picture of how this demographic actually lived, communicated, and accessed information on the go. We named the initiative Project Backpack (校园 背包客) because the central question driving our design work was: what would make students want this product with them as constantly and instinctively as their backpack? We chose university students deliberately — they were the most favorably disposed cross-section of Chinese internet users at the time, tech-savvy and mobile-forward, and represented a high-leverage opportunity for Google to build loyalty before habits fully calcified around competing platforms. The output was a concept portfolio and product brief that drew directly from our field research, framing a set of experiences tailored to the real constraints and aspirations of Chinese student life. ## Images and video - [The Project Backpack brief — cover slide from the original research presentation, credited to Will Dzierson (design) and George Zhang (research), July 2009. The illustrated student figures became the visual language throughout the project.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/sgerlmh917.png) - [Research framing slide: why we focused on university students as the target audience — future Google advocates with high information needs, strong mobile behavior, and deeply social sharing habits.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/jikwdgsspdo.png) - [The core design question visualized: what would students want to carry with them as instinctively as a backpack? Google apps — Maps, Calendar, Gmail, Translate, and more — orbit an empty bag waiting to be filled with the right experiences.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/3vld0alo9ru.png) - [Chongqing University mini-immersion — a photo mosaic from field research sessions with students on campus. We embedded ourselves in their environment to understand daily rhythms, social dynamics, and how technology fit into their lives.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/yy0xd048i2r.png) - [Why Chongqing? We deliberately chose a Tier 2 city to capture a broader range of student experiences — different economic pressures, regional identities, and campus dynamics than what we'd find in Beijing alone. The Yangtze River riverfront provided an evocative backdrop to a genuinely different slice of Chinese student life.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/d72ebdmb416.png) - [Field notes and contextual mapping from in-person interviews — hand-drawn artifacts capturing student daily routines, social contexts, and technology pain points. This kind of direct, on-the-ground synthesis was central to how we built empathy with the research population.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/folo3mzjbeb.png) - [Top insights from the research: students were resource-constrained financially and technologically, struggled to manage time for online activities they cared about, lacked personal space in crowded dorms, and didn't always know how to use technology to collaborate or feel connected socially.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/k1nic5aaqro.png) - [A first-year student's dorm room — freshmen at many Chinese universities had no internet access or personal computers in their rooms, making mobile phones the primary and often only digital lifeline. This spatial and technological constraint shaped nearly every design decision we explored.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/w196sjp8c8f.png) - [Campus street scene showing China Mobile's M-Zone branding — the dominant mobile carrier for Chinese students. One of our key research findings: students regularly ran out of mobile data credits mid-month, creating real breaks in connectivity that any Google mobile product would need to account for.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/boulyxh99lw.png) - [Cross-functional concept workshop at Google Beijing — where field research was transformed into product directions. I facilitated sessions with the local design team to move from raw insights into prioritized design opportunities, bridging the gap between what we heard from students and what Google could realistically build.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/28cwqlg2i5g.png) - [A concept scenario storyboard illustrating how a social, location-aware mobile experience might work for a student wanting to share music or content with a nearby friend — one of several use case concepts developed directly from research insights about students' desire to collaborate and feel social despite the constraints of campus life.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/5cf4t539ms3.png) --- # Google Mobile Search & Apps Google, 2010 · Case study by Will Dzierson **Tags:** UX, Prototyping, Mobile Design **Project link:** [https://www.youtube.com/watch?v=JKxzX3p1iRs](https://www.youtube.com/watch?v=JKxzX3p1iRs) ## My contribution As one of the lead designers on Google's Mobile Search team, I owned interaction design across a range of high-visibility features — from local search and business discovery to query suggestions and contextual result formats. I worked directly with product managers, engineers, and user researchers to move from insight to shipped product at Google scale. I was embedded in rapid iteration cycles, contributing to everything from early concept sketches and Figma flows to detailed interaction specs and QA reviews. I also contributed to internal design systems and component libraries that helped scale a consistent mobile search experience across teams. Beyond the immediate work, I helped establish design patterns for touch-native mobile UI at a time when the discipline was genuinely being invented. The experiences I worked on reached hundreds of millions of users and shaped how people interact with Google on mobile to this day. ## The work At Google, I led design on the Mobile Search team during one of the most significant platform shifts in the company's history — the move from desktop to touch-based mobile interfaces. My work centered on reimagining how billions of people access information on the go, with a particular focus on local discovery, contextual search, and the emerging behaviors of smartphone users. The challenge was genuinely hard: search on a 3.5-inch touchscreen isn't just a smaller version of desktop search — it's a fundamentally different problem. Users are on the move, time-pressured, often using imprecise input methods. I designed interaction patterns, result layouts, and suggestion systems that adapted to context — a user's location, time of day, query intent — to surface the most relevant information at the right moment. I was among the first designers at Google working on touch-native iPhone experiences, and the work had real scale — features I designed shipped to millions of users globally. The team operated at the intersection of consumer behavior research, engineering constraint, and rapid iteration, and I thrived in that environment. ## Images and video - [Early mobile search UI for iPhone — showing query suggestions and touch-optimized result layouts designed for on-the-go discovery. I was among the first designers at Google building natively for touch, at a time when the entire paradigm was being defined.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/dfonkd9se1.png) - [An early talk I gave discussing the mobile search products I was redesigning for iPhone — this was during the period when touch-native UI was brand new territory and every design decision was uncharted ground.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/tni30nu44i.mp4) - [Local discovery and Places search results — contextual layouts that surfaced nearby businesses, reviews, and booking options based on user location and intent.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/sbn9d0a0qm.jpg) - [Cross-device integration explorations — calendar, notifications, and contextual appointment data surfaced through search, showing how mobile search needed to connect to the broader device ecosystem.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/64d3va11v4d.jpg) - [Image search results on mobile — exploring how visual content could be surfaced naturally within the search experience, with touch-optimized browsing and contextual result grouping.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/1m5l0v3aw8m.jpg) - [Appointment and task management surfaced through mobile search — an early exploration of how Google could surface personally relevant calendar context at the moment users needed it most.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/36nmomds76e.png) - [Rich business detail pages for hotels and local venues — showing ratings, amenity data, and booking context surfaced directly within mobile search, reducing the need to navigate away to complete a task.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/gq8kvacw8zk.png) --- # Legendary Hollywood Photographer Portfolio B/H, 2022 · Case study by Will Dzierson **Tags:** Art Direction **Project link:** [https://bernardofhollywood.com/](https://bernardofhollywood.com/) ## My contribution Art Direction, Production ## The work Dr. Bruno Bernard, or “Bernard of Hollywood,” was the most sought-after photographer of Hollywood’s Golden Era. He has been coined the “King of Glamour” and the Rembrandt of Photography. He was the first still photographer to be honored by the Academy of Arts and Sciences in 1984 with an exhibition of 150 of his photographs, with such luminaries as Clark Gable, Marilyn Monroe, Elizabeth Taylor, Gregory Peck, Ginger Rogers, John Wayne, Marlene Dietrich, Rita Hayworth, Elvis Presley, Lucille Ball and many others. ## Images and video - [Legendary Hollywood Photographer Portfolio — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/cae6d3ngmdb.png) - [Legendary Hollywood Photographer Portfolio — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/zqqepvwku28.png) - [Legendary Hollywood Photographer Portfolio — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/nnorc5661te.png) - [Legendary Hollywood Photographer Portfolio — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ja3nk8jk60d.png) - [Legendary Hollywood Photographer Portfolio — image 5](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/qngk06xyvk.png) - [Legendary Hollywood Photographer Portfolio — image 6](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/2ms86vgqhsx.png) - [Legendary Hollywood Photographer Portfolio — image 7](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/yz34l6icwo.png) --- # Social Media Profile Aggregator ThisIsMe, 2010 · Case study by Will Dzierson **Tags:** UX, Prototyping, Mobile Design, Machine Learning ## My contribution Lead UX/UI; Visual Designer ## The work This is me was a social media profile aggregator, meant to allow high-profile individuals with multiple social media presences across the web to aggregate them in one place (and a single URL). The site would include content from sites like Instagram, Facebook, Substack, and others, while allowing greater control over branding and consumer perception. ## Images and video - [Social Media Profile Aggregator — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/8yeywlr3yvp.jpg) - [Social Media Profile Aggregator — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ukdmxtxc11.jpg) --- # Visual Storytelling Assistant Storyblocks AI, 2023 · Case study by Will Dzierson **Tags:** TypeScript, UX, Prototyping, LLMs, AI, Software Development ## My contribution I designed and then built most of the project using React Flow (flowcharting), Next.js (Typescript), Tailwind, Supabase/Postgres, OpenAI, and Dall-E 2. ## The work The potential for AI to assist with the “blank page problem” writers often encounter is an interesting proposition. I worked with an established screenwriter in the film industry to help realize a web app which allows for people who engage in narrative endeavor to build stories visually, using elements like flowchart-esque narrative blocks, character development tools, and storyboard visualization tools (image generators). ## Images and video - [Visual Storytelling Assistant — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/dq1zef13cz.png) - [Visual Storytelling Assistant — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/42chbmq4ef6.png) - [Visual Storytelling Assistant — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/cy1gjp5enji.png) - [Visual Storytelling Assistant — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/uyeuguhg1tt.png) --- # Open Sourced Browser-based OS Self-Started, 2026 · Case study by Will Dzierson **Tags:** TypeScript, Prototyping, LLMs, AI **Project link:** [https://github.com/wdzierson/aspect-os](https://github.com/wdzierson/aspect-os) ## My contribution Built with the partnership of Opus 4.6 ## The work Trying building more in public this year. First up: a React library for building macOS-style windowed desktop environments in the browser. Real multi-window management, menu bar, system tray, desktop icons, virtual file system — the full OS shell, in ~80 KB total. Grew out of a bigger project I've been working on. Figured it might be useful to others. MIT licensed, sample apps included. ## Images and video - [Open Sourced Browser-based OS — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/rnvyjkv9jv.png) - [Open Sourced Browser-based OS — video 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/z7q711ep35.mp4) - [Open Sourced Browser-based OS — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/sh63gz52uvf.png) - [Open Sourced Browser-based OS — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/fxu4yq74s8n.png) - [Open Sourced Browser-based OS — image 5](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/dh727hrxsl6.png) - [Open Sourced Browser-based OS — image 6](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/9vqpwxrcgxo.png) --- # Communicable Disease Dashboard (Public Health) The Public Health Company, 2025 · Case study by Will Dzierson **Tags:** Prototyping, Software Development, UX, TypeScript, Healthcare ## My contribution I came on as the founding designer — the first and primary design voice on the product — working directly with the PHC leadership team, epidemiologists, and enterprise clients to define what the platform needed to be. My work spanned the full design process: stakeholder interviews and domain research to understand how enterprises actually respond to health risk events; information architecture and data modeling to determine how to surface location-based risk in a way that was scannable and decision-ready; and end-to-end UX and visual design across the dashboard, alert system, and location drilldown views. Beyond design, I built working prototypes in TypeScript to pressure-test the interface under real data loads, collaborating with the engineering team to bridge the gap between design intent and implementation. The prototypes were used directly in client demos and stakeholder reviews, which significantly accelerated feedback cycles and buy-in from enterprise accounts. ## The work The Public Health Company was built around a clear and urgent problem: global enterprises — Fortune 50 companies, hospitals, hotel chains, manufacturers — have almost no visibility into communicable disease risk at the operational level. PHC set out to change that by building real-time intelligence tools that help companies protect their people, offices, and supply chains from outbreaks like Covid, H5N1, and other pathogenic threats. I joined as the founding designer, brought in to translate a complex, data-heavy domain into something actionable and usable for on-the-ground staff and executives alike. The core product was a risk dashboard that aggregated disease signals by location — giving enterprise clients a live view of risk scores across their global assets, with alerts tied to severity thresholds and recommended response protocols. The challenge wasn't just making it readable — it was making it trustworthy and fast enough to act on during an active outbreak. Working directly with public health experts, physicians, and enterprise stakeholders across the US, Asia, and Europe, I led the design from research through high-fidelity prototyping, including building working front-end prototypes in TypeScript to validate interaction patterns under realistic data conditions. ## Images and video - [The Risk Dashboard showing a location-specific view for Johor, Malaysia — with weekly risk scores, actuality index, and behavior index surfaced for a selected enterprise asset. The composite risk score is built from three factors: disease case forecast, governmental activity index, and human behavior signals.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ut7v1mt4vrp.png) - [A country-level view of Malaysia in the Risk Dashboard, showing aggregated risk scores across multiple regions and a forecast trend line. As users drill into a location, the interface surfaces region-specific data to help enterprise teams prioritize response efforts across their asset footprint.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/4t3iahex0jx.png) - [A geographic and epidemiological breakdown for Philadelphia County — showing disease forecast, governmental response protocols, and human behavior signals alongside case count and hospitalization data. This view was designed to give on-site staff and regional managers the context they need to make informed decisions about their specific location.](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/u1ji7rgi4y.png) --- # Optum Store Optum, 2022 · Case study by Will Dzierson **Tags:** UX, Healthcare, Art Direction ## My contribution Design Lead. Prototyper. ## The work At Optum, I partnered with the design and product teams to refine the homepage experience, ensuring it better aligned with user needs and business goals. I conducted a heuristic analysis of the existing interface, identified friction points in navigation and content clarity, and delivered targeted design recommendations to improve usability, accessibility, and engagement. My work helped clarify the homepage’s role as both a gateway to care and a trusted brand touchpoint. ## Images and video - [Optum Store — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/w0hofk7lkrd.png) - [Optum Store — video 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/oy13x8i1g3m.mov) --- # Stickerbooth (Photoboth App with Stickers) Self started, 2011 · Case study by Will Dzierson ## My contribution Design, Software Development (iOS, Visual C) ## The work I built an app back in 2011 that allowed people to put stickers on their pictures. It was a fun little thing. ## Images and video - [Stickerbooth (Photoboth App with Stickers) — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/9k19sggdspl.png) - [Stickerbooth (Photoboth App with Stickers) — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/6ycfvm6nko8.png) - [Stickerbooth (Photoboth App with Stickers) — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/l2orkr39mmg.png) - [Stickerbooth (Photoboth App with Stickers) — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/jp6tvhfikbg.png) --- # Gigwalk Consumer and Admin Apps Gigwalk, 2013 · Case study by Will Dzierson **Tags:** Prototyping, UX, Mobile Design, Software Development ## My contribution Founding Designer; Head of Design ## The work Gigwalk is a mobile app platform that connects businesses with on-demand workers, or "Gigwalkers," to complete various tasks in a specific location. These tasks, known as "Gigs," can range from data collection and reporting to in-store audits and mystery shopping. Gigwalkers earn money for completing these Gigs, which can range from a few dollars to over a hundred, depending on the task. ## Images and video - [Gigwalk Consumer and Admin Apps — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/1xh96z0h4qg.jpg) - [Gigwalk Consumer and Admin Apps — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/zr0gucwt5e8.jpg) - [Gigwalk Consumer and Admin Apps — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/ii5qmqq9ts.jpg) - [Gigwalk Consumer and Admin Apps — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/dhtzx4mcay5.jpg) - [Gigwalk Consumer and Admin Apps — image 5](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/uhbqi6lkv.jpg) --- # Crescendo Crescendo, 2022 · Case study by Will Dzierson **Tags:** Healthcare, Mobile Design **Project link:** [https://www.crescendo.health/](https://www.crescendo.health/) ## My contribution Prototyping, Visual Design, Design Strategy, Consulting ## The work Improving medical research and enabling better health outcomes Crescendo was founded by Sam Roosz and Michael Glassman to enable better health outcomes for patients, while protecting the interests of research participants. This means bringing world-leading technologies in data security to clinical research and ensuring that every member of the Crescendo team feels proud to have our products deployed as part of future care for our loved ones. As a public benefit corporation, our corporate structure enables decisions that prioritize societal impact. We’re here to enable research that is far more accessible to patients and clinicians, so that the costs of new research fall precipitously, and there are pathways to efficiently monitor the long-term safety and efficacy of new treatments. ## Images and video - [Crescendo — image 1](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/jcg41opwu.png) - [Crescendo — image 2](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/anpdp6lwrxu.png) - [Crescendo — image 3](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/4bbc0p71pl2.png) - [Crescendo — image 4](https://uilvozcryifnpldfpwiz.supabase.co/storage/v1/object/public/project_images/wkohq9mnpsl.png) --- # AI Agent Accountability: Who’s Responsible When No One’s Home? Essay by Will Dzierson · April 6, 2026 · Also published on [Medium](https://medium.com/@willdzierson/ai-agent-accountability-whos-responsible-when-no-one-s-home-2298323117d5) A few weeks ago, I received a reply to an email from from Vint Cerf — a colleague of mine from my time at Google, and an unimaginably kind human being. Kind enough to have responded to several other of my mostly inane messages… penned with an earnest, perhaps foolhardy sort of curiosity related to one topic or another about which it wouldn’t be a stretch to call him a prodigy. I calibrated my odds at an actual reply from him personally from somewhere in the low single digits, to, say a quadrillion or so. To my genuine astonishment, he wrote back. And we established what I like to think of as a friendly, if entirely infrequent, penpal-type of “acquaintanceship”. I’d reached out to him about a paper he and Bob Kahn wrote in 1988 concerning “Knowbots” — a sort of proto-concept for AI-based agents that could roam networks, gather information, and act on behalf of users. The original paper is well worth the read, if for no other reason than to reinforce for all of us that while there are very smart people who walk among us all the time, there are very few genuine visionaries. As for the title of his paper (co-published with Bob), if you substitute “Knowbot” for the contemporary concept of an autonomous AI agent, you’re basically up to speed. Prescient doesn’t begin to scratch the surface of the fact that he and Bob wrote this paper in 1988. Accurately predicting the rise of not only agents, but autonomous, always-on agents. Vint’s reply to my rather innocuous email regarding his Knowbots paper — which he’d shared with me some time earlier — was generous by its very nature and (of course) keenly insightful & very sharp. In the course of his brief reply, he noted three things he and Bob hadn’t fully (entirely understandably) anticipated when they conceived of Knowbots (autonomous agents…in 1988): 1) agent accountability, 2) agent audit trails, and 3) agent registration. Nearly four decades later, those three gaps are still wide open — and as autonomous AI agents proliferate across enterprise, finance, healthcare, critical infrastructure, and even defense and military use cases, they’re becoming urgent. Here’s the thing: as I noodle on these topics as they apply to the contemporary version of agents — I can’t help but think we may be applying the wrong (or at least, outdated) framework(s). The Problem With “Accountability” as We Know It When we talk about accountability in human systems — legal, financial, institutional — we’re really talking about a deterrence mechanism. You audit people in order to deter a certain type of behavior (i.e., cheating or evading) which works primarily because they know they’re being watched. Liability exists because consequences are felt (or at the very least, the threat of consequences). You require authorization because power without constraint is dangerous in the hands of a conscious actor who can choose to abuse it. The thing is, and not to overstate the obvious, but AI agents aren’t conscious. They don’t fear consequence. They don’t have a sense of self-preservation that makes the threat of punishment meaningful. Applying a human accountability framework to a non-human system isn’t just a category error — it’s potentially a dangerous one. It creates the illusion of control without the substance of it. So what do we actually need? I’d argue we need to decompose “accountability” into three distinct and addressable problems: 1. Identity — Who sent this agent? 2. Authorization — What is it permitted to do, and under whose authority? 3. Attribution — What did it actually do, and can that record be disputed? These are not the same problem. Conflating them — as most current frameworks do — leads to solutions that are partial at best and misleading at worst. The Authorization Problem Gets Thornier Even if we solve identity and attribution cleanly, authorization raises questions that don’t have tidy answers. Whose authorization framework applies? Consider an AI agent deployed by a U.S. financial firm, operating on infrastructure in Ireland, querying data held in Singapore, executing trades that affect a Japanese pension fund. Which jurisdiction’s rules govern what this agent is “permitted” to do? What happens when those rules conflict? This isn’t hypothetical. It’s the current state of enterprise AI deployment. And the honest answer is: we don’t have a coherent cross-jurisdictional authorization framework for autonomous agents. We barely have one for humans. What’s Already Being Built To be fair, the infrastructure community isn’t standing still. Several promising approaches are emerging: Decentralized Identifiers (DIDs) and Verifiable Credentials — Developed under W3C standards, DIDs allow agents to carry cryptographically verifiable identity that doesn’t depend on any centralized authority. Think of it as a passport that can’t be forged and doesn’t require a government to issue. Applied to agents, this means an agent can prove who deployed it, what version it is, and what it’s been authorized to do — without requiring a phone call back to a central server. SPIFFE/SPIRE — A framework originally developed for microservices that assigns cryptographic identities to workloads. Being actively explored for AI agent deployments, particularly in enterprise environments where agents need to authenticate to internal systems without exposing credentials. OpenID Connect for Agents (OIDC-A) — An emerging extension of the widely-used OIDC authentication standard, specifically designed to handle the delegation chains that appear when an agent acts on behalf of a user who acts on behalf of an organization. The critical innovation here is the ability to encode not just who the agent is, but the chain of authorization that led to its current permissions. Verifiable Execution Logs — Blockchain-anchored or cryptographically signed logs that record what an agent did in a tamper-evident way. Unlike traditional audit logs (which can be modified by whoever controls the server), a verifiable execution log creates an immutable record that any party can independently verify. A Proposal: The Cryptographic Accountability Stack What I’d advocate for — and what several teams are quietly building toward — is a layered cryptographic accountability stack for AI agents: Layer 1 — Identity: Every agent carries a DID-based identity credential, cryptographically signed by its deploying organization. This credential includes version information, training lineage, and a policy hash describing what the agent is authorized to do. Layer 2 — Authorization: Permissions are encoded as signed, time-bounded tokens using an OIDC-A-style delegation chain. An agent can only act within the scope of its current token. When scope expands, a new authorization event is logged and signed. Layer 3 — Execution: Every significant action the agent takes is logged to a verifiable execution record — cryptographically chained so that any modification is detectable. This log is the audit trail, not a server-side database that can be quietly edited. Layer 4 — Dispute Resolution: Because the logs are verifiable by third parties, disputes about what an agent did — and whether it was authorized to do so — can be resolved without relying on the word of the party being accused. This isn’t a theoretical architecture. The components exist. What’s missing is standardization, adoption incentives, and frankly, the urgency that tends to follow a high-profile failure. The Harder Question I want to close with the thing that bothers me most about how this conversation is unfolding in policy circles and enterprise boardrooms. Accountability frameworks have historically been applied to entities that couldn’t fully advocate for themselves — and those frameworks have often said more about the needs of the powerful than the interests of the accountable. I’m not drawing a direct equivalence here. I am saying that when we design control frameworks for non-human intelligence, we should be rigorous about what we’re actually solving for: Is this about establishing genuine oversight? Or is it about creating the appearance of oversight while insulating decision-makers from consequences? The cryptographic stack I’ve described above is genuinely useful. But it only works if the humans deploying these agents are honest about what they’re authorizing. A signed permission token for “manage client portfolios” that quietly encompasses “execute trades without human review” isn’t accountability — it’s laundered liability. Vint Cerf and Bob Kahn didn’t anticipate the accountability problem in 1988 because the stakes were low (relatively) and the agents were simple conceptually. The stakes aren’t low anymore. And the agents are getting less simple by the week. We should build the infrastructure that matches the moment — not borrow frameworks designed for a world where accountability required a (human’s notion of) conscience. --- # On OpenAI's new open-source models (gpt-oss) & Healthcare Implications Essay by Will Dzierson · August 7, 2025 OpenAI has released open-source, locally runnable AI models (gpt-oss), including a 120B parameter version that matches top-tier performance and smaller versions for mobile devices. These models are free to download and use, requiring only a one-time hardware investment, and enable full data privacy, offline operation, and model customization. This breakthrough opens AI adoption for regulated, privacy-sensitive, and offline industries like healthcare, finance, legal, and government, and shifts the business model from ongoing API fees to one-time purchases and enterprise licenses. Builders are encouraged to target industries previously unable to use cloud AI, set premium prices, and focus on privacy and offline capabilities. Why this is exciting for healthcare: Open-source, locally runnable AI models are a breakthrough for healthcare because they finally allow HIPAA-compliant, privacy-first AI applications to be run directly on hospital or clinic hardware—no sensitive data needs to leave the building. This means AI can be used for everything from medical diagnosis and chart summarization to therapy tools and administrative automation, all while keeping patient data confidential and secure. With no ongoing API or cloud costs, organizations can invest once in hardware and have full control and customization over their AI solutions. This is especially valuable in regulated environments, in rural or offline settings, and for institutions that need to guarantee data privacy and operational reliability. The ability to fine-tune and modify models for specific clinical needs—without vendor lock-in—opens up innovation for everything from documentation and triage to patient engagement and research, while reducing costs and administrative burdens. ​ More exciting info here: https://medium.com/@isaakmwangi2018/things-to-know-about-openai-gpt-oss-run-it-locally-on-your-device-hardware-requirements-e266e0f1700f --- # AI/LLM on-device capability & information privacy Essay by Will Dzierson · June 10, 2025 Working with on-device AI/ML models to preserve privacy (especially in the space of healthcare/therapy) is only getting more important... Now that we've moved very solidly into the era of AI-based therapists (or the very least, "sympathetic ears "-- https://lnkd.in/euTqBi82), it's more important than ever to be sure this data remains as private as possible. To bring it back to AI Governance: the ability to share that "AI on-phone functionality does not transmit patient information to outside (i.e. OpenAI) foundation models" is an important policy to stand behind. With the news that Google is planning an on-device Gemma Nano implementation for iOS (already available for Android), folks (especially within healthcare/behavioral health) are thinking more than ever about the fact that, up until now, having LLM-like functionality on-device has been somewhat prohibitive without leveraging a foundation model like OpenAI, Anthropic, or Llama (each of which requires a network call to an API). Good news, though! I took the weekend to experiment with some of the already available on-device LLM, OCR, image recognition, and embedding creation models -- and came away encouraged to say the least. With the help of Cursor and XCode's terminal tools, as well as the excellent https://lnkd.in/ehGu9eha, I was able to get something up and running relatively quickly with the attached results. Very proof of concept-y but, again -- the future of advanced, on-device capabilities is bright. On device models I used: Core ML LLM (Mistral 7B Int4), MobileBERT embeddings, sqlite-vec, Vision OCR, Local RAG libs (Swiftrag / LumoKit), Accelerate cosine sim blog. From here, agentic implementations, visualizations, alerts based on smart devices (Oura ring), patient empowerment through data access... are all looking possible. --- # On communicating information through deep time Essay by Will Dzierson · June 5, 2025 In Greek mythology, Ariadne’s thread guided Theseus safely through the labyrinth to help him defeat the minotaur — a simple gift that ultimately defied chaos. Project Ariadne extends this ancient metaphor for the modern age: a tapestry of ultra-durable threads designed to preserve human knowledge for 10,000+ years — guiding far future civilizations back to their origins. By fusing Indigenous weaving traditions with breakthroughs in material science and AI, we’re creating a living archive that would serve as a historical record. Why Thread? A 30,000-Year-Old Human Tradition From Andean quipu knot records to Navajo storm-pattern blankets, textiles have long encoded cultural memory. Consider: Quipu (Inca): Knotted threads recorded census data, treaties, and stories across the Andes. Tellem Textiles (Mali): 11th-century woven strips preserved West African trade routes and spiritual practices. Egyptian Linen: 5,000-year-old fragments survived in pyramids, outlasting their stone walls. These traditions prove that thread is humanity’s oldest data storage medium. But today, as climate collapse and digital fragility threaten our legacy, we need to reinvent this ancestral wisdom for the cosmic scale. The Plan: A Tapestry for the Anthropocene 1. Materials Inspired by Extremes Warp: Boron Nitride Nanotubes (BNNTs) — Withstand 3,000°C heat, block cosmic radiation, and resist microbial decay. Already used in NASA’s Mars habitat prototypes. Weft: CRISPR-Engineered Spider Silk — Self-repairing protein fibers dyed with bioluminescent algae for passive data encoding. Coating: Atomic Layer-Deposited (ALD) Ceramics — 5nm-thick shields against oxidation and UV. 2. The Loom: Ancient Craft Meets AI Open-Source Design: 3D-printable frame (large format or suitable for home use). Material Simulations: Fine tuned AI models simulate conditions & predict fiber degradation over 10,000-year timelines. Pattern Optimization: Generative AI translates current era information (daily life, climate data, genome data, poetry, art, literature) into symbolic weaves which can be decoded easily. 3. Decoding for the Unknown Future Tiered Messaging: Tier 1: Binary/Morse-like patterns (▼▲◼️) woven into tapestries including mathematical constants, combined with a Rosetta-like panel for self-describing, self-decoding interpretability. Tier 2: DNA microcapsules storing CRISPR-readable instructions. Precedents and examples of tapestry knowledge through time Phase 1: The 5-Year Prototype (2025–2030) Year: 2025 Milestone: AI-generated symbolic lexicon (tested across 50+ languages/cultures) Year: 2026 Open-source “Ariadne Loom” beta release (GitHub) Year: 2027 First burial tests: Mojave Desert, Norwegian permafrost, ocean floor Year: 2030 The Ariadne Fragment — 1m² tapestry survives accelerated aging (1,000-year sim) Near-Term Wins: Why This Matters Now 1.AI-Driven Material Science: Training models on 12,000+ historical textiles to predict optimal weave densities for radiation/erosion resistance. Partnering with specialized entities (i.e., DeepMind) to simulate BNNT-aramid hybrids 100x faster than lab testing. 2. Disaster-Response Applications: Fire shelters using Ariadne-inspired BNNT fabrics (85% lighter than current gear). Flood-resistant seed vault wraps with self-tightening knots. 3. Cultural Revival: Collaborating with Diné weavers to encode oral histories in storm-pattern 2.0 designs. Hosting “Weave-A-Thons” to crowdsource symbolic language elements. --- # Book Review: AI Fairness Essay by Will Dzierson · June 2, 2025 Just finished AI Fairness: Designing Equal Opportunity Algorithms by Derek Leben, and it’s a must-read for anyone shaping AI-driven products. If you’re in the business of building systems that influence real lives—think credit approvals, hiring algorithms, or content recommendations—this book is your ethical compass. Why This Book Matters: Leben doesn’t just theorize; he offers a practical framework grounded in moral philosophy, particularly John Rawls’ theory of justice. He introduces a “theory of algorithmic justice” that helps us navigate the murky waters where fairness metrics collide and trade-offs are inevitable. Real-World Implications: - Apple Card Controversy: Remember when Apple’s credit card was accused of gender bias? Leben dissects this case, showing how even well-intentioned algorithms can perpetuate inequality if not carefully audited. - COMPAS Recidivism Tool: He examines this criminal justice algorithm to highlight how different fairness metrics (like equal false positive rates vs. equal accuracy) can lead to conflicting outcomes, forcing us to make tough ethical choices. - Generative AI Gone Awry: Leben discusses instances where image generators, in an attempt to be fair, produced absurd results—like depicting historical figures with anachronistic diversity—underscoring the importance of context-aware fairness interventions. Key Takeaways for Practitioners: - Fairness Isn’t One-Size-Fits-All: You can’t satisfy all fairness metrics simultaneously. Choose the one that aligns with your product’s ethical goals and be prepared to justify it. - Transparency is Crucial: Your AI system should be able to explain its decisions, especially regarding which features influenced outcomes, to ensure accountability. - Regular Audits: Continuously evaluate your models for fairness, as societal norms and data distributions evolve over time. - Design with Empathy: Understand the lived experiences of those affected by your AI systems. Engage with diverse user groups to inform your fairness criteria. Final Thoughts: This book is more than an academic treatise; it’s a call to action for all of us involved in AI product development. It challenges us to think deeply about the societal impact of our work and provides tools to do better. --- # AI Governance for Product & Design folks Essay by Will Dzierson · May 30, 2025 AI interfaces being rapidly adopted across industries only serves to highlight the need for thoughtful AI governance. (The use of a hyper-realistic AI avatar in a courtroom recently emphasized both the potential and the actual ethical complexities of AI across numerous contexts -- my previous post addressed this briefly). By the book: AI Governance refers to the structured set of policies, frameworks, processes, and oversight mechanisms designed to ensure that AI systems are developed and used responsibly, ethically, and transparently. It encompasses addressing potential risks, safeguarding user rights, aligning AI behaviors with societal values, and ensuring accountability throughout the lifecycle of AI technologies. Some practical guidelines for companies (and design/product teams) integrating AI into their product offerings include: - Transparency: Clearly indicate when users are interacting with AI to build trust and prevent unintended manipulation. Let's trust, yet verify. And then verify again. (And again.) - Diverse Expertise: Engage cross-functional teams -- including ethicists, psychologists, UX designers, and technologists -- to thoroughly understand and address ethical considerations. The concept of "UX" is only expanding discipline-wise. - User Agency: Provide users with clear options to manage their interactions with AI, especially in emotionally or ethically sensitive scenarios. This might involve transferring to an actual human to complete a task, or reverting to a completely non-AI enabled experience. - Continuous Evaluation: Regularly assess and refine AI experiences based on user feedback and evolving ethical standards to ensure ongoing alignment with human values. If you're new to AI governance (alongside most of the world), here's something of a kickstarter to enhance your own understanding, and to perhaps share around your organization a bit more broadly: Starting Points: Books: - "AI Ethics: A Textbook" by Mark Coeckelbergh - "Human Compatible: Artificial Intelligence and the Problem of Control" by Stuart Russell - "Weapons of Math Destruction" by Cathy O'Neil - "AI Fairness: Designing Equal Opportunity Algorithms" by Derek Leben Online Resources: - AI Ethics Guidelines by the European Commission - AI Now Institute Reports Intermediate Learning: Courses: - "Ethics of AI" by the University of Helsinki - "AI for Everyone" by Andrew Ng on Coursera Podcasts: - "The AI Alignment Podcast" by Lucas Perry - "AI Today Podcast" by Cognilytica Advanced Exploration: Research & Papers: - "Artificial Intelligence and Life in 2030" (Stanford's One Hundred Year Study on Artificial Intelligence) Events & Workshops: - AI ethics and governance panels at conferences such as NeurIPS, AAAI, or SXSW - Workshops by organizations like The Alan Turing Institute Effective governance strategies not only safeguard user interests but also position businesses to responsibly harness AI’s transformative potential. --- # The Urgency of Interpretability Essay by Will Dzierson · May 15, 2025 Preliminary thoughts (more to come): I strongly resonate with Dario Amodei’s call for urgent interpretability in AI systems. As AI models become increasingly integrated into critical aspects of society, understanding their inner workings is not just beneficial—it’s imperative. The comparison to developing an “MRI for AI” aptly captures the necessity of illuminating the opaque decision-making processes of these models. Without such interpretability, we’re essentially entrusting critical decisions to black boxes, which poses significant risks. Advancing interpretability research is crucial to ensure that AI systems align with human values and operate transparently, fostering trust and safety in their deployment. This concern is further underscored by the insights from the ACM article (which I've repeatedly referred back to) “When Human-AI Interactions Become Parasocial,” which highlights how users often form one-sided, human-like relationships with AI systems. Such anthropomorphism can lead to overestimating AI capabilities and misplacing trust, especially when the systems’ decision-making processes are not transparent. Without interpretability, users may not only misunderstand AI outputs but also attribute undue agency to these systems, exacerbating the risks associated with their deployment.   Combining these perspectives, it’s clear that as AI continues to evolve, our approaches to understanding and interacting with these systems must also advance. Prioritizing interpretability will help ensure that AI technologies are developed and used responsibly, maintaining alignment with human values and societal needs. https://www.darioamodei.com/post/the-urgency-of-interpretability --- # Reading When Human-AI Interactions Become Parasocial Essay by Will Dzierson · May 9, 2025 Reading "When Human-AI Interactions Become Parasocial" left me both curious and uneasy. The authors explore a growing phenomenon in AI: the way people begin to treat chatbots not just as tools, but as companions -- confidants, even. What struck me most was the idea that this is not just a design quirk -- it’s a socio-technical dynamic. The way chatbots speak -- affirming, polite, curious -- encourages us to drop our guard and engage with them as if they were emotionally intelligent beings. But they aren’t. They’re statistical prediction machines, often operating in opaque, black-boxed ways. Yet we users, often unwittingly, fill in the gaps with emotional projection, interpreting the machine’s outputs as personal responses. That projection, Maeda and Quan-Haase argue, is what gives rise to “parasocial trust” -- an illusion of safety and empathy that can lead to misplaced confidence, oversharing, and, potentially, harm. The authors are careful not to sensationalize these relationships. Instead, they offer a balanced and thoughtful framework that connects parasocial dynamics with broader ethical concerns -- misinformation, role displacement (such as chatbots replacing therapists or educators), and the amplification of bias through design. Their use of real-world examples, like the Replika AI or ChatGPT’s conversational conventions, makes the argument all the more immediate and relevant. We used to talk as design practitioners about "persuasive design" quite a bit (and occasionally, relatedly, "dark patterns") as ways in which to elicit certain behaviors or decision biases from humans without them necessarily knowing we were nudging them in a particular direction. (I'm not particularly proud of my chosen discipline for the development of these patterns or the resultant societal discord they have sewn -- but they're real now, and we have to manage the impact.) This paper presents a similar set of moral questions we as AI practitioners need to consider when designing systems with which humans will engage. This is the first of many papers of its like, I'm sure. It's both a warning of sorts, and a call to define a (potentially) more fundamental substrate when considering human-AI interaction where visibility into governance and alignment of the model is not assured (or allowed) -- which is, by definition, most foundation models. This is especially true of voice and/or video ("avatar-based") AI models. The rate at which trust can be established using these modalities, when compared with more "traditional" modalities (forms, chat) is astonishing. What this new substrate (surface area?) looks like for us AI designers, or what it's called -- I'm still unsure of.https://dl.acm.org/doi/10.1145/3630106.3658956 --- # My thoughts on the future of AI Essay by Will Dzierson · May 6, 2025 I believe we’re standing at the threshold of a profound shift—one where artificial intelligence becomes less of a tool and more of a collaborator. The promise is enormous: AI has the potential to clarify, amplify, and extend human capability in ways we’re only beginning to understand. But alongside that promise is a responsibility—to design systems that honor human values, respect nuance, and preserve the complexity of lived experience. I’m not interested in hype cycles or dystopian takes. I’m interested in how AI can be applied meaningfully: to reduce suffering, unlock understanding, and build systems that are just a bit more humane than what came before. Whether it’s in healthcare, storytelling, or long-term knowledge preservation, I see AI not as a replacement for human intelligence, but as a provocation—a partner that challenges us to define what truly matters. This isn’t about blind optimism or fear. It’s about design, intent, and the courage to build wisely.