Viewing profile — dexterlagan

I consider using LLMs to generate code akin to metaprogramming. We created these programs to write programs, which, if it was seen as such, would be considered the highest form of …

That's pretty crazy. I didn't know they rewrote that much of CC. Thanks for this. Hard to trust anything in that space anymore. I tend to have some trust in the model, so far Claud…

Hmmmm. OK. So yesterday I was accepting all these prompts in the manual mode for a solid hour, then I switched to Auto. Side project, nothing critical (creating ComfyUI workflows w…

Had this song in my head, wrote some lyrics, hummed it into Suno, generated the song. I love listening to it. What a world. I make little tools like that all the time. My GitHub is…

"I want to be careful here, because “taste” is doing a lot of work and it is easy to hear it as decoration." Claude, is that you? I haven't seen a human write like this, ever, and …

Do you have examples we could try?

I'd say competition is good. The large investments in datacenters in the US is also a sign that both politicians and business owners understand the risk of falling behind China, an…

I was about to write this. So refreshing, after having looked at hundreds of semi or completely AI-written articles. Even the news look like sterile Claude now. This is so human.

Love this. Maki.sh looks solid, will try!

Terrible writing. Uses em dashes incorrectly and for no good reason. I like them too, and I'm saddened that they're now seen as a sign of AI slop. But that writeup is sadder.

Article ends with "Get in touch with me (LinkedIn, email) if you want to get some details about a product we built aiming at giving control back to the engineers.". Already then.

Wild(er) idea: expect future AI to maintain current AI-written code.

For those who aren't aware, Cursor sports one of the best LLM harnesses for coding. The app itself is annoying to use compared to their CLI counterparts, but the harness is widely …

I thought that was what everybody was doing all day since LLMs came out. I certainly spend most of my free time asking questions, pulling refs, parsing research papers and brainsto…

There's a reason why this tech is called disruptive. The same phenomenon can be observed on Reddit. You'll see a lot of knee-jerk reactions to anything that looks AI, as in 'Thanks…

If you have suggestions for good Discord servers, please share. The bullshittery is coming out of my ears. I don't quite know where to turn to anymore. There's HN. Reddit is gettin…

Anthropic has since found and fixed the perf. problem: https://www.anthropic.com/engineering/april-23-postmortem

Many of us tested 27B and 35B side by side, and the dense model is significantly smarter. It indeed is slower, but 35B makes a lot of mistakes 27B doesn't.

Bonus points for MD being readable even when it's not parsed. More bonus points for Sublime Text displaying it in plain text and still looking great. Good enough++

Same as top comment, have spent a lot of time on local models. IMHO, qwen3.5 is the very first model that is actually usable for serious work, ever - and I've tried them all. The 3…

I used to use LLMs to 'clean up' my own writings, and in the end I agree with the author here: it doesn't really help. The reader will have this impression of 'too perfect', and wi…

The tech debt this title speaks of only applies if humans have to deal with it. Tech debt is an assumption made on the grounds that humans are still programming and AI does not evo…

There is one thing everybody forgets when making such predictions: companies don't stand still. Nvidia and every other tech business is constantly exploring new options, taking ove…

Execution is cheap? How about you try a video game, and not 3 obvious and worthless automations I could have made as a quick fix at lunch time.

I had the same idea. I think this is very useful. As it is it does look like a proof-of-concept, and that's OK. I'd develop this as a book recommendation site and simply link to th…