Anthropic released Cowork a few days ago. I tried it for a few minutes and it was unusable for me. It crashed several times and felt very buggy. What I liked was the idea of the productivity plugin. I set up my own version of it that is agent agnostic and can be easily customized. Here are the main ideas and a link to a template to fork.
I had similar thoughts recently. I wouldn't consider myself "the thinker", but I simply missed learning by failure. You almost don't fail anymore using AI. If something fails, it feels like it's not your fault but the AI messed up. Sometimes I even get angry at the AI for failing, not at myself. I don't have a solution either, but I came up with a guideline on when and how to use AI that has helped me to still enjoy learning. I'm not trying to advertise my blog and you don't need to read it, the important part is the diagram at the end of "Learning & Failure": https://sattlerjoshua.com/writing/2026-02-01-thoughts-on-ai-.... In summary, when something is important and long-term, I heavily invest into understanding and use an approach that maximizes understanding over speed. Not sure if you can translate it 100% to your situation but maybe it helps to have some kind of guideline, when to spend more time thinking instead of directly using and AI to get to the solution.
Some years ago, I was at a conference and attended a very interesting talk. I don't remember the title of the talk, but what stuck with me was: "It's no longer the big beating the small, but the fast beating the slow". This talk was before all the AI hype. Working at a big company myself, I think this has never been more true. I think the question is, how to stay fast.
And, to add to that, how to know when to slow down. Also, having worked at a big company myself, I think the question shifts towards "how to get fast" without compromising security, compliance etc.
Great story, thanks for sharing. Besides the part where it says "Other people will see its glory and join their smaller snowballs into it.", it sounds a bit like marriage too.
Very interesting, thanks for sharing this. After reading Karpathy's recent tweet about "A few random notes from claude coding quite [...]" it got me thinking a lot about offloading thinking and more specifically failure. Failure is important for learning. When I use AI and they make mistakes, I often tend to blame the AI and offload the failure. I think this post explores similar thoughts, without talking much about failure. It will be interesting to see the long-term effects.
Wrote down some thoughts on AI-assisted coding after reading Karpathy's "A few random notes from claude coding [...]". Mostly about the ego part, what happens to learning when you offload failure, and which skills matter more now. Hopefully, sharing this will help others stay positive despite all the hype and negative headlines. Thanks for reading :)
I would like to understand how he creates the transparent/blurry terminal overlays in front of the video. My hypothesis is that he creates two videos. First, the screen recording with alpha channels. Second, the video of him coding. He then layers them on top of each other. Any other ideas? Which screen recording software supports this kind of recording?
Hint: I think that he is not streaming the video as a wallpaper, because in some videos you can see his screen with the regular wallpaper.
I'm actually not so interested in the editing software. I think that part is doable with any modern video editing software. I'm more interested in the screen capture and how to preserve the transparency. He mentioned Cleanshot X somewhere for screen recording, but I doubt that he is using it for the recording. I tried it and it did not support the transparency that you can see in his videos.
reply