You don't need to learn every AI tool out there

I'm very active in AI communities, and everyone seems to have a very specific answer. Use this model. Use this skill. Use this framework. The thing I keep running into is that everybody is right and everybody is wrong at the same time. This one is me thinking out loud in the car about why that happens, and what I think you should actually do about it. The short version is something I've been saying for years, long before AI. The best workout is the workout you'll actually do. The best diet is the diet you'll actually follow. The best tool is the tool you will actually use. There is no method, no skill and no model that is one size fits all, and I think that's fine. It's the part most people leave out when they tell you what to use. Something I've come to appreciate is that you don't have to adopt a tool just because it's the best of a few options. You can open Claude and describe the thing you want, and five minutes later you have it. That changes what "the right tool" even means. So if you're out there trying to learn AI and comparing yourself to people who seem to be doing well, I don't think you're going to get the same results by copying their stack. When you pick up something new, just ask whether it moves you further along than you were yesterday. If the answer is yes, adopt it. If it isn't, you haven't missed anything. This is the first time I've recorded one of these. If you like the format, let me know down below and I'll keep doing them. 0:00 Everyone in AI thinks they have the answer 1:22 Why nothing is ever black and white 2:31 Everybody is right and everybody is wrong 2:56 Cherish the tool you'll actually use 4:26 You can just build your own tools 5:27 The only question worth asking 5:54 First time rambling in the car