Artificial Intelligence

I’m not 100% certain I love this article. I made myself write it in less than one hour. It’s core point is supposed to be around written language being the dominant skill going forward which replaces today’s primary skill of knowing how to do stuff (visual skill). I will likely revisit this topic in the future as I think about it more. For at least the next 5-10 years, professional work is going to become more language-based, not less. We are entering a period where being able to express yourself clearly in writing will be more critical than ever. ...

I had a little fun a few days ago having AI do some forecasting for this weekend DC winter weather event. Figured now that it’s Thursday night I would give the AI chatbots a chance to create a new forecast as we have more updated models. Below is the prompt and the results. I’ve updated my prompt a bunch. I’m really interested in continuing to refine the prompt and using it more frequently. Ultimately I would love to create a process of having it forecast and then tracking accuracy of the forecasts over time. Project for another day! ...

Will it snow? So this is just a quick post on some fun I had this evening with three of the major AI chatbots: Claude, ChatGPT, and Gemini. I’m in the Northern Virginia area and there is a potential winter storm coming this weekend. It has been getting some wild forecast ranges, from maybe a dusting to the Apple Weather app currently showing ~20 inches. So lets just say no one really knows what is going to happen, which is basically true of every winter storm in the DC area. ...

Why LLM’s will always make mistakes and we shouldn’t call them hallucinations I saw a tweet from Paul Graham a while back about how as LLMs become better their hallucinations will become more convincing. And it makes sense, a smart confident person saying something wrong often sounds more reliable than a less confident person saying the right thing timidly. Even more so, as you get good answers from the smart confident person you become more trustworthy and are less likely to question and double-check their future answers. That’s both a reality of and defect of human thinking. ...

I’ve been using GitHub Copilot, Claude, and ChatGPT to help me code for a couple of years now. But AI has always been in the assistant role – I was the one in the driver’s seat, asking technical questions or letting GitHub Copilot generate sections of code based on what I was writing. It makes you a more productive developer because it’s faster than Googling, but at the end of the day, I was still the one fundamentally coding, just with better tooling. ...