kenkahn.bsky.social @kenkahn.bsky.social · 29/09/2026It was a challenging and long (5 days) effort to add voice input to ToonTalk 3D. Enabling free form speech to control a rich interface requires a frontier LLM. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 29/09/2026On the 17th of October I'll be co-hosting a Logo Foundation online workshop entitled "In the Age of AI, Is Programming Still a Good Activity for Children?" el.media.mit.edu/logo-foundat... 000
kenkahn.bsky.social @kenkahn.bsky.social · 27/09/2026I did more experiments with three LLMs pretending to be Seymour Papert commenting on some of my recent efforts. I found only Claude’s imitation of Papert to be insightful and thought-provoking. docs.google.com/document/d/1...docs.google.comMore experiments with LLMs imitating Seymour PapertMore experiments with LLMs imitating Seymour Papert - home page 27 September 2026 I recently posted . It included some experiments with AIs roleplaying Seymour Papert. I decided to do more experime... 000
kenkahn.bsky.social @kenkahn.bsky.social · 24/09/2026Seymour Papert wrote “Is Programming a Good Activity for Children?” in 1993. I thought one of his examples would be a good test of the latest versions of Claude and ChatGPT. They did very well. docs.google.com/document/d/1... 101
kenkahn.bsky.social @kenkahn.bsky.social · 21/09/2026I’ve been busy creating a full-featured 3D ToonTalk with Claude for the last few weeks. It supports images, sounds, puzzles, 3D models, interactive lessons, meta-programming, and anima-gadgets. 15,000 new lines of code since the first release. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 15/09/2026My 5-year old granddaughter asked Claude Fable for a game, played it, suggested changes, and repeated ten more times. It went very well. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 30/08/2026In the midst of making many improvements to ToonTalk 3D, I decided to talk to a Seymour Papert bot about my efforts. I found my ideas being clarified and I gained insights into my research. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 27/08/2026After reading about a study of how well AI models are at writing turtle graphics programs that reproduce a given sketch, I wondered how good Claude would be at this. Its attempts were impressive even if most were only approximately correct. docs.google.com/document/d/1... 100
kenkahn.bsky.social @kenkahn.bsky.social · 24/08/2026While creating ToonTalk 3D, Claude claimed it couldn’t create an artifact version of the app. Then I remembered a solution from a month ago. After creating the new version I added some new features including improvements to viewing numbers. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 22/08/2026Together with Claude Code I created a 3D first-person implementation of ToonTalk. Several innovations emerged. The over 10,000 line program took a week involving almost 150 exchanges. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 16/08/2026Constructionist toolkits are usually focused on STEAM topics (math, science, and art). With Claude’s help I created an AI-powered civics app around the challenges of describing rules and values. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 13/08/2026I’ve been quiet lately since I’ve been busy using Claude Code to port ToonTalk’s 210,000 lines of C++ that I began writing in 1992 to the web. After extensive testing and my 525 messages to Claude we finally succeeded. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 01/08/2026After receiving feedback from the creator of Comic Chat I forwarded it to Claude. Claude updated the app. I then asked Claude to provide the option of using Gemini Nano which is built into Chrome. docs.google.com/document/d/1... 011
kenkahn.bsky.social @kenkahn.bsky.social · 30/07/2026It sure was easy for Claude to make the Comic Chat app I finished yesterday multi-user. And to optionally support AI from ChatGPT, Claude, or Gemini. docs.google.com/document/d/1... 010
kenkahn.bsky.social @kenkahn.bsky.social · 29/07/2026I asked the top coding agents from Anthropic and OpenAI to port Microsoft’s Comic Chat to the web. Claude did well, so then I asked it to add AI responses. GPT Sol 5.6 didn’t do so well. In the following screenshot only Anna was human generated. docs.google.com/document/d/1... 010
kenkahn.bsky.social @kenkahn.bsky.social · 27/07/2026I asked ChatGPT 5.6 to look over the child-friendly app for mechanistic interpretability of language models I recently created with Claude. It came up with and implemented several good improvements. docs.google.com/document/d/1... 100
kenkahn.bsky.social @kenkahn.bsky.social · 25/07/2026The good and bad news is that today’s AI can create decent games with just a few sentences. It empowers kids to play the video games they invent. But the learning opportunities are very impoverished. docs.google.com/document/d/1...docs.google.comClaude and ChatGPT agents design and implement very different unicorn gamesClaude and GPT agents design and implement very different unicorn games - home page 25 July 2026 Curious about how good the latest language models are at generating code with nearly no guidance,... 010
kenkahn.bsky.social @kenkahn.bsky.social · 19/07/2026I discussed the Atlantic essay “What if It’s Not the Phones?” with Claude, then asked ChatGPT to fact check it. This led to an interesting number of rounds as they argued about the research around phone and social media bans. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 17/07/2026I asked Claude for an essay on the history of AI with a focus on the role of the Perceptrons book and common (mis)understandings of “good old-fashioned AI”. ChatGPT had several criticisms of the essay and I copied and pasted between them until they were happy. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 16/07/2026Last week Anthropic released a paper introducing a new way to inspect and modify the thoughts of LLMs. Claude suggested we build a kid-friendly web app that reproduces this research. It turned out very well. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 10/07/2026Together with Claude I made a very large web app where a student can train a language model and apply a variety of mechanistic interpretability tools. The app includes an AI guide that reacts to events, gives tours, and answers questions surprisingly well. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 05/07/2026I noticed a discrepancy in the Nature article “Generative artificial intelligence creates delicious, sustainable, and nutritious burgers” and discussed it with Claude. It decided on its own to find and run the code to get to the bottom of the matter. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 04/07/2026With Claude’s help I added AI guides to three more apps, one of which I’ve never used before. In the process Claude gained enough experience to generate resources to improve the process of adding guides. docs.google.com/document/d/1...docs.google.comAdding AI guides to three apps and helping Claude to learn to do this betterAdding AI guides to three apps and helping Claude to learn to do this better - home page 4 July 2026 Adding an AI guide to the Logo turtle program interpolator After being pleased with how addin... 000
kenkahn.bsky.social @kenkahn.bsky.social · 01/07/2026I asked Claude to add an AI guide to a 2700-line app. As the enhanced app is run the guide explains and highlights sections. The user can ask questions and steer to the demo. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 30/06/2026Claude helped me build an app that given two Logo turtle programs creates a program that can interpolate (and extrapolate) between them. In 2D or 3D. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 28/06/2026When I saw 2 butterflies dance and then fly apart, I had questions. Thinking of Seymour Papert’s 30-year old vision of a knowledge machine, I instigated a conversation with Claude. docs.google.com/document/d/1...docs.google.comFrom Seymour Papert’s Sleeping Giraffes to my Looping ButterfliesFrom Seymour Papert’s Sleeping Giraffes to my Looping Butterflies - home page 28 June 2026 In 1993 Seymour Papert published The Children’s Machine. Starting on page 6 he tells this story: [The ... 000
kenkahn.bsky.social @kenkahn.bsky.social · 28/06/2026Rather than search through 60 Google Docs that needed updates to some links I asked Claude Cowork to do it for me. A short prompt was all I needed. docs.google.com/document/d/1...docs.google.comOne short prompt to Claude Cowork saved me an hour of tedious editingOne short prompt to Claude Cowork saved me an hour of tedious editing - home page 27 June 2026 The problem This year I’ve occasionally asked a chatbot to create design and plan documents that I the... 000
kenkahn.bsky.social @kenkahn.bsky.social · 24/06/2026Claude and I designed a game which views language as a microworld, a place where learners grow a construction kit. Communication is facilitated by defining words in terms of the few words the aliens know. There is a purely verbal version. docs.google.com/document/d/1...docs.google.comA game that treats language as a construction kitA game that treats language as a construction kit - home page 24 June 2026 Claude and I designed a game which views language as a microworld, a place where learners use and grow a construction kit... 000
kenkahn.bsky.social @kenkahn.bsky.social · 19/06/2026With Claude’s help I made a customized version of Claude. docs.google.com/document/d/1... A test I ran was comparing the process of co-creating an app with the personalized and default versions of Claude. The apps they created fuse together different LLM responses. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 16/06/2026Everyone should tinker with AI. I explored asking ChatGPT to act like a journalist who is interviewing me. Perhaps, if I persisted, it would have resulted in a decent article but it didn’t work very well. But it was worth trying. docs.google.com/document/d/1...docs.google.comAn LLM wasn’t much help in recording my stories about Marvin MinskyAn LLM wasn’t much help in recording my stories about Marvin Minsky - home page 16 June 2026 A friend who liked my stories about Marvin Minsky suggested I write them down. I thought it would be in... 000
kenkahn.bsky.social @kenkahn.bsky.social · 13/06/2026I made an app for children to design and experiment with agentic systems that learn in a constructionist manner. Students can construct agents that create images, poetry, and short stories. Claude did much of the design and all of the implementation. docs.google.com/document/d/1... 100
kenkahn.bsky.social @kenkahn.bsky.social · 06/06/2026Twenty years ago I made a “metagame” for students to construct space games like Lunar Lander. With the help of both Codex and Claude Code I reconstructed the program. I then created a fresh design and implementation with ChatGPT. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 01/06/2026Computers can use reinforcement learning (RL) to learn how to play games. I asked Claude to create an app for children where they can explore RL on Tic-tac-toe, Connect Four, and Pong. I’m impressed how well this turned out. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 30/05/2026Claude and I made a kid-friendly app for kids to explore the evolution of little Logo programs. It supports both artificial and natural selection. It includes an ask-the-AI feature that works very well. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 29/05/2026I asked Claude to make a modern version of the 1979 TT2500 variant of Logo by adding color, pen width, and more. I programmed an animated aquarium in it. I worked with ChatGPT to create a 2200-line Python program to add music to my 1979 MANDALA program docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 24/05/2026You CAN code. Even young children can have a conversation with a chatbot and end up with a web app that does what they want. I’ve been posting about this for 3 years and wrote a book on the topic. WIRED used to be different. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 23/05/2026I challenged Gemini Antigravity to resurrect my programs, emails, and letters from almost 50 years ago. It re-implemented all but one of my old programs. Claude completed the set. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 21/05/2026I just read news.stanford.edu/stories/2026... from Stanford. 000
kenkahn.bsky.social @kenkahn.bsky.social · 20/05/2026With Claude I created a large app (7500 lines of code) for exploring symbiosis in evolution. Sadly, I got negative, uninteresting, results. docs.google.com/document/d/1...docs.google.comClaude helped me build a Symbiotic Micro-Behavior Evolution Lab that didn’t evolve anything very interestingClaude helped me build a Symbiotic Micro-Behavior Evolution Lab that didn’t evolve anything very interesting - home page 20 May 2026 Making the app This exploration began after watching the excel... 110
kenkahn.bsky.social @kenkahn.bsky.social · 18/05/2026With the help of Codex I created a simulation of the 1918 Influenza Pandemic along with learning materials and an ask-AI feature. In 2012 creating the simulation took me almost half a year, this new model took me half a day. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 16/05/2026I made several enhancements to the local model chatbot app I created with Gemini last month. But Gemini let me down when I tried to add more models and features. I succeeded after switching to Claude. docs.google.com/document/d/1...docs.google.comUsing a chatbot to create an offline chatbot for vibe codingUsing a chatbot to create an offline chatbot for vibe coding - home page 7 April 2026 Updated and enhanced 16 May 2026 When I read about Google’s release of open Gemma 4 models, I thought it w... 010
kenkahn.bsky.social @kenkahn.bsky.social · 14/05/2026I continued my discussion with Claude after reading Corporations Constituting Intelligence. Besides getting a better understanding of AI governance issues, I was surprised to see so many introspective statements from Claude. docs.google.com/document/d/1...docs.google.comContinuation of my discussion with Claude after reading Corporations Constituting IntelligenceContinuation of my discussion with Claude after reading Corporations Constituting Intelligence - home page 14 May 2026 After I read Abiri’s March 2026 paper Corporations Constituting Intelligence I ... 000
kenkahn.bsky.social @kenkahn.bsky.social · 13/05/2026I just read Public Constitutional AI, a paper by a Yale Law professor. I had lots of ideas and questions and then proceeded to have a very good, deep, conversation with Claude about how to govern AI. docs.google.com/document/d/1...docs.google.comHaving a long, in-depth, conversation with Claude about a Public Constitutional AI paper led to a deeper understandingHaving a long, in-depth, conversation with Claude about a Public Constitutional AI paper led to a deeper understanding - home page 13 May 2026 After reading PUBLIC CONSTITUTIONAL AI by Gilad Abiri... 000
kenkahn.bsky.social @kenkahn.bsky.social · 11/05/2026After reading “A Lo-Fi Rebellion Against AI” in the New Yorker I wondered how hard it would be to prompt ChatGPT and Gemini to generate images that don’t have an AI look. Chatbots thought many of the images might be mistakenly thought to be by people. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 10/05/2026I played with a new tool from Anthropic that let me (dimly) see the inner thoughts of an LLM as I asked it to write poetry, create an app, and answer questions about my book. docs.google.com/document/d/1... 010
kenkahn.bsky.social @kenkahn.bsky.social · 07/05/2026I scanned a paper I wrote over 40 years ago and asked Codex to make an HTML version of the paper where the reader can run and modify all the examples in the paper. I’m happy with the result. docs.google.com/document/d/1...docs.google.comTurning a 40-year old paper into an interactive document with the help of CodexTurning a 40-year old paper into an interactive document with the help of Codex - home page 7 May 2026 I’m a big fan of active essays (now known as explorable explanations). After reading a paper ... 000
kenkahn.bsky.social @kenkahn.bsky.social · 05/05/2026I returned to the app I co-created with ChatGPT a few weeks ago where the Logo turtles draw on 3D shapes. I added cubes and several new features. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 02/05/2026With Gemini’s help I built an app that periodically creates illustrations in real time based upon what is being said. youtu.be/96jN2OCOfLs?... 200
kenkahn.bsky.social @kenkahn.bsky.social · 27/04/2026Claude helped me build an animated message to aliens that begins with turtle graphics and ends with science and mathematics. Inspired by the 1960 book on the Lincos message. docs.google.com/document/d/1... 000
kenkahn.bsky.social @kenkahn.bsky.social · 26/04/2026I asked ChatGPT whether a hack that enables programs to use the OpenAI API without a costly API key could be done in web pages as well. API usage would then count against the model usage quota of the ChatGPT subscriber. It is possible but very clumsy and hacky. docs.google.com/document/d/1...docs.google.comCreating web apps that use OpenAI’s API as a subscriberCreating web apps that use OpenAI’s API as a subscriber - home page 26 April 2026 When I read about a project that uses the OpenAI GPT API without an API key, I wondered if it was possible to creat... 010
