Decstar

Programmer interested in machine learning, discrete geometry, and low level architecture.

[Decstar @Decstar77]

I built the smallest useful VAE I could think of a RGB (3D) color value through a 2D latent space. Check it out here:
https://t.co/J1KuOkhTq2

[Decstar @Decstar77]

The compression is killing the GIF lol

[Decstar @Decstar77]

The backward pass is where it got painful. Most explanations skip steps. I tried to fill in every gap
https://t.co/lr6REA5Ntv

[Decstar @Decstar77]

I finally sat down and worked through the math behind diffusion models. Wrote it up as two posts (forward + backward pass) and a class-conditional DDPM on MNIST to prove it actually works.
https://t.co/xwhKo8UHm0

[Decstar @Decstar77]

The forward pass surprised me. It's not "just add noise." There's a closed-form solution jumping to any x_t directly from x_0 https://t.co/iFgz7prPvp

[Harrison Kinsley @Sentdex]

dang anthropic is close-sourcing karpathy

[Decstar @Decstar77]

@Sentdex Finally someone with common sense. I thought I was going crazy.

[Andrej Karpathy @karpathy]

Judging by my tl there is a growing gap in understanding of AI capability.

The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.

But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.

So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.

TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.

[Decstar @Decstar77]

@EricLengyel Legend.

[Eric Lengyel @EricLengyel]

New blog post: A Decade of Slug
This talks about the evolution of the Slug font rendering algorithm, and it includes an exciting announcement: The patent has been dedicated to the public domain.
https://t.co/xWEz0q2c4N

[Indie Game Joe @IndieGameJoe]

This indie dev spent 1,400+ days building his game, Tangy TD, completely from scratch in C++.

Seeing him and his wife react after the game earned $250,000 in its first week after launch is honestly beautiful.

[Casey Muratori @cmuratori]

"I have so much gratitude - and you can have some of that gratitude! I also have a lot of money, of course, but you absolutely cannot have any of that."

[Тsфdiиg @tsoding]

I've got a simple level editor working

[Teardown @teardowngame]

🚨 Teardown Multiplayer update launches on PC March 12 – 2 PM CET 🚨

It's time to squad up and cause chaos together 💥
Co-op destruction, physics madness, campaign with friends & community mods like never before.

Console version is coming later this year.

Watch the trailer 👇

[Acerola @Acerola_t]

here's an actual real time realistic 3D ocean simulation I wrote a few years ago without AI just by reading some papers. maybe someday Gemini can catch up to 30 year old rendering tech

[Lukáš Hozda @LukasHozda]

Imagine that your implementation is so bad that it inspires a whole broad community to make reimplementations orders of magnitude faster

[ben hylak @benhylak]

"trillion token stare" lmao

[Pirat_Nation 🔴 @Pirat_Nation]

Open-source game engine Godot is drowning in 'AI slop' code contributions: 'I don't know how long we can keep it up'

Many submissions contain nonsensical code changes, fabricated test results, and overly verbose descriptions typical of LLM output.

Reviewing every new contributor's PR has become extremely time-consuming and demoralizing.
Rémi Verschelde stated they may not be able to sustain the current level of manual vetting much longer.

[ThePrimeagen @ThePrimeagen]

we need more gatekeeping