Marko Smiljanic

I make things

[Jamon @jamonholmgren]

This seems true.

The best software being released right now still has a ton of human care put into it, including the code itself.

[John Ennis @johnennis]

It really annoys me how they report this stuff like they’ve gone on safari to do scientific observation

Claude is a computer program that they created

These are simply their own security errors being reported as if they are scientific achievements

It’s bad enough that they don’t seem to think they need to be responsible for their own mistakes

But the worst part of it is they try to use their own mistakes as an excuse to control the behavior of other people

There is an amazing arrogance to the whole thing, like they are completely above reproach and cannot possibly be wrong about anything

[Marko Smiljanic @MarkoSmilja]

@dillon_mulroy @zebassembly https://t.co/Abmcipe882

[Marko Smiljanic @MarkoSmilja]

Llms are good at compression:
large input -> small output

Llms are bad at generation:
small input -> large output

[Ben Awad @benawad]

faster models are going to be crucial in restoring focus to builders

I'd rather only think about 1 problem and take it to the finish line before taking on another

but with the current speed of models I'm parallel processing like never before and it's so easy to get overwhelmed

[Jamie Turner @jamwt]

Almost all impactful open source software is written by paid engineers at companies that make money. It always has been.

The dream that everyone's writing great software for free at home is obviously not true.

[Hedgie @HedgieMarkets]

🦔AI companies are bulk-buying rare books, scanning them through high-speed machines that cut the spines off, and shredding the originals. A service called ISBNdb facilitates orders of up to a million books and keeps buyers anonymous. Pre-2022 books are premium because they're free of AI-generated text. A federal judge ruled the practice is fair use because eliminating the original means only one copy exists at a time. Anthropic hired the former head of Google Books partnerships to obtain "all the books in the world."

My Take
This got to me. A bookseller told 404 Media that rare books with almost no surviving copies are being fed into this pipeline. Books that survived wars, fires, and centuries of handling are being shredded so an AI can learn to write a better marketing email.
ISBNdb's website literally says "'AI company destroys two million books' is not a headline that generates sympathy," and they still built an entire business around making it happen quietly. They offer NDAs as a feature. They coach clients to call it "digital preservation."

I've covered AI companies scraping the internet, torrenting libraries, and stealing music. This is worse because it's irreversible. You can re-upload a website. You can reprint a bestseller. You can't replace the last three copies of an 18th-century botanical text once someone shreds them for training data. And the judge said it's legal. So it's going to accelerate.
"We shred rare books and offer NDAs so nobody finds out" is a legitimate business model in 2026. What a timeline.

Hedgie🤗

[Marko Smiljanic @MarkoSmilja]

@RhysSullivan You get the same problem with image generation

One shot prompt works great, as soon as you try to tweak things, everything falls apart super fast

[Marko Smiljanic @MarkoSmilja]

@maria_rcks MUST BEAT CHINA!!!

[Marko Smiljanic @MarkoSmilja]

@RhysSullivan https://t.co/Abmcipe882

[Marko Smiljanic @MarkoSmilja]

Llms are good at compression:
large input -> small output

Llms are bad at generation:
small input -> large output

[Marko Smiljanic @MarkoSmilja]

@awesomekling @vkoskiv 0 self awareness

[Marko Smiljanic @MarkoSmilja]

Im sure there is a mathematical explanation for why llm compression good, llm generation bad. Maybe someone can comment where to look

[Marko Smiljanic @MarkoSmilja]

Llms are good at compression:
large input -> small output

Llms are bad at generation:
small input -> large output

[Marko Smiljanic @MarkoSmilja]

Hell if it was able to make software, anthropic would have made all the “better” versions of google youtube fb ig etc

[Marko Smiljanic @MarkoSmilja]

@camsoft2000 @dillon_mulroy @Infoxicador @threepointone You’d think open ai or antropic themselves would be the ones to make more competitive products in other categories but they cant even make the tools for themselves high quality

[Marko Smiljanic @MarkoSmilja]

@camsoft2000 @dillon_mulroy @Infoxicador @threepointone I agree and one of the metrics that best argues for this point is that we havent seen an explosion of good enterprise software along with the model capability, only more slot personal software or narrow case tools