Results for
I’ve installed Claude-code, MATLAB MCP Core Server, and now Puppeteer on my MacBook Pro. Puppeteer can navigate and operate web pages like Perplexity Comet or the new Claude Chrome Extension. The new wrinkle is MATLAB in the loop.
Claude-code and MATLAB MCP installation are described at Experiments with Claude code and MATLAB MPC Core Server . To install and configure Puppeteer, I used Claude App and its ability to use my MATLAB’s access to system files. The installation includes a Google Chrome for testing browser that is independent of (and does not interfere with) my normal Chrome browser. Puppeteer installation took just minutes of my approving various steps, and quitting and relaunchinbg Claude App. A minor hiccough was overwriting a special fetch connector configuration but that was readily fixed. The resulting linkage is Claude (cloud) ↔ Claude Desktop App ↔ MCP Server (local) ↔ Puppeteer ↔ Chrome for Testing (local) as well as the link to MATLAB on my laptop.
As a very first test, I selected Wikipedia from the Claude App suggestions. We navigated to a page Lorenz system in the Chrome for testing browser where Claude dismissed a prompt for donations to Wikipedia. ( I was like “What is Anthropic’s valuation and why didn’t you donate?” but said nothing.) and extracted content and summarized, many of the equations beautifully formated. I issued the following prompt: “Take a look at the differential equations there and at the example solutions and their parameter values. Then create a MATLAB script in my folder MATLAB/ClaudePuppeteer to reproduce the illustrations and run the script.” After clicking to approve various steps, presto.
The screen shot below shows 1) Claude App (upper left) after the process completed, 2) a MacOS Finder window showing the Lorentz Attractor.m in a folder ClaudePuppeteer that Claude had previously created for me to test Puppeteer functionality, 3) the Lorentz system wiki page in the Google Chrome for testing browser (lower right), 4) the LorentzAttractor script open in MATLAB (upper right), and 5) various figures created by the MATLAB Script.
Um, wow!
MatGPT was launched on March 22, 2023 and I am amazed at how many times it has been downloaded since then - close to 16,000 downloads in one year. When AI Chat Playground came out on MATLAB Central, I thought surely that people will stop using MatGPT. Boy I was wrong.
In early 2023 I was playing with the new shiny toy called ChatGPT like everyone else but instead of having it tell me jokes or haiku, I wanted to know how I can use it on MATLAB, and I started collecting the prompts that worked. Someone suggested I should turn that into an app, and MatGPT was born with help from other colleagues.
Here is the question - what should I do with it now? Some people suggested I could add other LLMs like Gemini or Claude, but I am more interested in learning how people actually use it.
If you are a MatGPT user, do you mind sharing how you use the app?
First, I felt that the three answers provided by a user in this thread might have been generated by AI. How do you think?
Second, I found that "Responsible usage of generative AI tools, such as ChatGPT, is allowed in MATLAB Answers."
If the answers are indeed AI generated, then the user didn't do "clearly indicating when AI generated content is incorporated".
That leads to my question that how do we enforce the guideline.
I am not against using AI for answers but in this case, I felt the answering text is mentioning all the relevant words but missing the point. For novice users who are seeking answers, this would be misleading and waste of time.
The MATLAB AI Chat Playground is now open to the whole community! Answer questions, write first draft MATLAB code, and generate examples of common functions with natural language.
The playground features a chat panel next to a lightweight MATLAB code editor. Use the chat panel to enter natural language prompts to return explanations and code. You can keep chatting with the AI to refine the results or make changes to the output.
Give it a try, provide feedback on the output, and check back often as we make improvements to the model and overall experience.
