aizk

Well, Gemini is just a bad model in general, you have two variables going on.

This is a rare instance where feeding this groundbreaking information into an LLM gives _them_ psychosis. I fed this to claude code and watched it verify the result in 7 different ways to be 100% certain, and it was just flabbergasted. Quite remarkable.

I want to see the Collatz Conjecture next!

Give the llms a few years, they'll be smart enough to make progress on that.

Who cares they're gonna discontinue it 6 months from now anyway.

Thinking of the ww2 plane with holes meme right now.

You're absolutely right!

A fun, simple game created by a friend of mine where you have to see if you can identify AI writing.

I use Sparky Fitness for all my fitness / health tracking. So, so happy I have everything stored locally, where I can do whatever I want with the data, and I know it's mine and it's private. Very easily to interface with your favorite llm of choice. https://github.com/CodeWithCJ/SparkyFitness/

Completely sensible legislation in my eyes.

That vibe coded app is generating billions. Does the word "vibe coded" mean anything anymore?

Funny you mention ORMs, I'm building a project with bun, and just using raw bun sqlite until I feel the app gets too complicated and I need it. AIs are really damn good at SQL, I can just trust them with it, and it keeps the project much lighter. A few years ago this would just sound stupid, but here we are.

There just isn't a demand for it. Have you seen just how dedicated nintendo hackers are? Check out decomp.dev https://decomp.dev/

Sam didn't lie, they are in fact a non profit.

I'm calling that this will be a dud. Price will be too high, it'll just be a watered down version of mythos, and just look at the track record of Anthropic's last few releases.

You may enjoy this - Claude autonomously grew a tomato plant: https://autoncorp.com/biodome/

Can you share that on GitHub?

What do you mean? Everyone is talking about Mythos.

Why do we store books in libraries, or write history textbooks? It's about preservation (and curiosity too).

The decomp dev guys are doing amazing work. It's also super educational too, if you're someone like me who's in just doing relatively simple AI / python / typescript work and rarely has to think about memory, hardware constraints, all that, it's a completely different world. Also, AI is finally getting to the point where it can do very difficult decompilation work, which is super exciting to me.

Kinda disagree. The code I had worked with was super unoptimized and difficult to run because there was just enormous amounts tribal knowledge that was just gone over the years as the company evolved. Also basically all the original devs had left. LLMs can help with that so much they know random minute details of lets say ADOdb and whatnot.

My first exposure to professional programming was writing VBA and SQL (yes, together) at a massive manufacturing facility that had really old equipment. Now with AI it's much easier to replace the code but VBA still has a stranglehold on legacy systems.

And now, the hackers will be able to scan the repos for more vulnerabilities. Vulnerabilities all the way down.

AI news and ESPN feels interchangeable sometimes.

I'm so glad Simon is documenting this. The field is evolving so fast, so rapidly, so hungry for data and money, that few are willing to zoom out and document everything big picture so we can see the changes over time. I mean do you guys remember "Do anything now"? Just a distant memory, a funny party trick.

I'm not sure why gh hasn't already implemented stricter measures / filters / tools for PRs. It would cut down on spam and also help save their servers that can't handle the increased AI load!

Back when I worked as a civil engineer I had a coworker named Joe. He was... I think 78. Poor guy didn't save up enough for retirement so he worked a bit part time. He knew the ins and outs of the field better than anyone, but had no idea how to use a computer (he marked up drawings and I put them into CAD). He mentioned this movie to me as AI (gpt) had just become a thing saying "it'll scare the hell out of you", and he recommended I watched it - I'm glad I did! Great guy, always told funny stories - "I was not a great dad but I was a damn wonderful grandpa!"

I might not necessarily agree with the haste / stability of this, but I commend Jarred for pushing boundaries on what AI coding is capable of, can't deny that. 4 years ago this would've seemed like science fiction.

Very slowly, and then, all at once...

How powerful will Opus become before they decide to not release it publicly like Mythos?