Two things I read last week that reminded me that not all people are not using generative AI and LLMs in the same way.
The first thing that I read was Timonthy Gowers‘s blog about math and LLMs. In the post, What sort of maths are LLMs good at?, he is writing about how AI is becoming very good at solving advanced math problems. However, when he talks about his own experiences with using gen AI, he seems to be talking about using chatbots like chatGPT, not agentic AI, which is what is how the math is being done.
The second thing I read was this: If you haven’t recently used Claude Code*, you might not understand where AI is at. It does a good job of showing how agentic AI is a different beast than the gen AI most people are used to.
If you are mainly using chatGPT or Gemini or CoPilot and want to know more about agents and agentic AI, then I recommend this primer from my employer, What is Agentic AI? | IBM. Or try this piece. You’ll see that what people like myself and others are doing with agentic AI to analyze documentation, design systems, write code, and much more.
Chatbots are still useful for getting information. But agentic AI is taking the use of AI to a new level. It’s worth learning about.
Smart People I Know 