HNHacker News
TopNewBestAskShowJobs

Reubend

1,992 karma · joined October 5, 2019

submissionscomments
Reubend··on The last time my family was replaced by technology
The internet has made it easier than ever to learn new skills. You don't need college unless you're targeting a career that requires, or expects, a degree.
Reubend··on Gravity Linux Alpha Release: Linux on the M4 Mac Mini with GPU and DCP Support
Very cool! I'd love to see a prominent section on their website explaining what does work and what doesn't work at the moment. I'm assuming there are probably quite a few things which aren't functional.
Reubend··on Tinfield 1 is an open weight coding model from Nigeria that beats Opus 4.8
It's a Qwen fine tune that's benchmaxxing a few sets of question types, and it's obviously nowhere near beating Opus 3.8 in real usage. Nothing to see here.

Normally, I'd give them a pass because at least it's open source, but then I read

> Raising pre-seed. So we welcome all angels and vcs

Which basically gives away that they're trying to fool investors into giving them money by exaggerating their results.

Reubend··on I built non-autoregressive decision models with RL a year ago
Oh, I missed that! So the Doom demo was potentially just "harnessmaxxing"?
Reubend··on Executive order on H-1B nonimmigrant visa administration
> ...the Secretary of Homeland Security shall take into account in any labor condition application... whether the employer sponsor directly or indirectly engaged in layoffs within the previous year or plans future layoffs...

Unfortunately, as a 2nd order effect, I wonder if we'll see fewer layoffs and more outright firing to balance it out.

Reubend··on Vial contains a new drug called PAC-3310
Wow, he did this all himself? Very impressive, and I'd love to see what comes of this. Does he have any plans for how he would fund clinical trials in the event that the preliminary results are promising enough for that?
Reubend··on On the Navier–Stokes Millennium Prize Problem
It's great that important discoveries like this can now routinely be accompanies by formalized proofs. The fact that it's being released alongside a Lean proof from Day 1, rather than the Lean proof being released months or years later, is super helpful for verifying that it's correct.
Reubend··on Tesla killing Solar Roof is leaving installers with six-figure losses
Is there any competitor in the USA? The idea of roof tiles which are themselves solar panels seems sensible to me, and it's more aesthetically pleasing than putting panels on top of normal tiles.
Reubend··on Launch HN: RonanRX (YC S26) – Personalized Peptides and GLP-1s
You manufacture the drugs yourself? That's impressive, and it feels like an interesting company.

Your website sucks though. We can tell you vibe coded it quickly and didn't put much thought into it.

Reubend··on AI Isn't Outthinking Mathematicians. It's Out-Remembering Them
For that particular example, it's now been proved that Mochizuki's proof is incomplete: https://zeli.app/en/story/48963019

I guess whether he will eventually fix those gaps and resolve the issues remains to be seen.

Reubend··on WorldClaw Agentic 3D open-world generation at scale
The assets look good, but for some reason the terrain looks too cartoonish to me.
Reubend··on [dead]
This article is from May of 2025.
Reubend··on Incentives Are for Losers
I think you've written a good counterargument, but I disagree with your claim that "morality is really just yet another incentive". Incentives must be external.
Reubend··on Seedance 2.5
It's just you. Any director, storyteller, or filmmaker can see the potential here to realize scenes that simply wouldn't be possible otherwise. But indeed there's plenty of slop coming out too.
Reubend··on DeepSeek-V4-Flash Update
They're always very understated in their update descriptions. This is actually a HUGE improvement in the model's capabilities rather than just a small tweak.
Reubend··on Our position on open-weights models
This seems hypocritical, out of touch, and hollow. "Safety testing" is just a hair away from censorship when it comes to government control.
Reubend··on Show HN: Optimize and serve models with Fable quality at half the cost
Yeah, this is just slop. No benchmarks, no concrete case studies, just some vibecoded "platform" to finetune models on your own traces.

Which is an idea that has some value, but also some weaknesses. And this implementation of it isn't forthcoming with that concept. You have to really dig in to understand what they're even talking about.

Reubend··on ARC-AGI Leaderboard
ARC-AGI3 doesn't seem like a great benchmark to me in the first place. It assumes a lot of human like tendencies which an AI either shouldn't or wouldn't have. Particularly in the genre of "gameplay" where unspoken assumptions from prior games inform our understanding of rules.
Reubend··on Anduril in talks to raise funding at $100B valuation, over 3x last year's mark
You've got a good point, but they might spend a lot on the software side too, and any kind of weapons manufacturing in America will be expensive compared to other countries.
Reubend··on Mage-Flow: Efficient Native-Resolution Foundation Model for Image Generation
There aren't many local image editing models, so this is a welcome contribution! It's very compact considering the good demo results from their webpage.
Reubend··on eSIM was supposed to replace SIM cards, but carriers turned it into a trap
My biggest gripe against eSIMs is that switching from one phone to another is difficult. I wish I could simply send a file to the other phone and get along with things.
Reubend··on Show HN: misa77 - a codec that decodes 2x faster than LZ4 (at better ratios)
Is there a way to regain some compression ratio in Snappy if you're okay with slow compression but need fast decompression?
Reubend··on Inkling: Our Open-Weights Model
Seems like this is particularly good at instruction following, but not as strong at coding as others. It's always great to get more diversity of open weight models though! I'll need to test this out to see what its "personality" is like.
Reubend··on Solving 20 Erdős Problems with 20 Codex Accounts Running in Parallel
Isn't the pursuit of knowledge alone good enough?
Reubend··on Stellantis to sell small Fiat Topolino EV for $13,995 in U.S.
I think there's some hidden demand for a small cheap car, but 19 MPH is indeed too slow. It needs to be able to go at least to slow highway speeds.
Reubend··on Show HN: Shirei, cross-platform GUI framework in native Go
> What is it that matters for "immediate mode"? Is it that the UI renders everything every frame? No. It's that you build the UI by describing what it should look like everyframe, based only (or mostly) on the data. This is why React won...

I don't think that's why React got so popular. React popularized unidirectional data flow, which is different than immediate mode rendering. This readme file seems to conflate the two of those.

Now that I think of it, couldn't one argue that React itself is a retained mode UI, since it choses which components to re-render and which not to?

Reubend··on Show HN: Ant – A JavaScript runtime and ecosystem
Can this engine be embedded into other programs?
Reubend··on Command and Conquer Generals natively ported to macOS, iPhone, iPad using Fable
> rendering DirectX 8 → DXVK → Vulkan → MoltenVK → Metal

Am I reading that right? It makes API calls that go through 5 different layers before actually getting rendered? That's kind of crazy. I'm surprised it works, although I guess the underlying libraries are solid enough that it shouldn't be unexpected.

Reubend··on Odin, Wikipedia and engagement farming
I'm sorry, but I agree with the wiki editors in this case. Odin is obscure. The author of this blog post seems to think it's well known, but I don't think that's substantiated.
Reubend··on DSpark: Speculative decoding accelerates LLM inference [pdf]
That understates how difficult it is to get to the level of performance they attained. The fact that it surpasses DeepSeek v4 in most ways shows that they accomplished some great work in this space.
Page 1 of 17Next →