Every week someone tells you there’s a new AI tool you must try or you’ll be left behind. New model, tool, workflow or approach. Miss it – and you’re out. This sounds familiar, because we’ve seen this pattern before.

Cool enthusiasts exist, they are early birds on adopting things. They are also content creators (or content creators just retranslate their thoughts in some way). But this whole situation reminds me of productivity porn and its successor – second brain apologists. 99% of content is similar and aimed to better deal with making content or courses to sell you this content, not resolving your real life or work problems. Second brain approach is good for researchers – people who professionally deal with large volumes of information, but not for everyone. And again – new creators are trying to convince us this new thing is better than the previous one. And we should try it. And buy some course.

In AI reality it’s getting easier and easier to join. New good tools appear, skill libraries get curated, the common knowledge base is expanding and at the same time becoming easier to use. Prompts are becoming less important now, because agents handle them better than most users can. Prompt engineering – goodbye (imho it was a silly direction from the start).

Comparisons of models – total miss. They change so quickly, and for typical tasks the gap between top models is not important. If something truly stands out, you will hear about it from every corner – like it was with Opus 4.5. For most users there is no need to follow benchmarks.

Do not forget about marketing and hype about AI. It’s a thing worth hype, but nowadays any hype is hyperbolized and FOMO is desired result because it makes you try and buy new stuff.

Use what’s popular, keep in mind the constant change, but do not make it a center of your life. By the time you finish that AI course, half of it will be outdated anyway.