Hacker News: supermdguy

New comment by supermdguy in "We ran AI agent swarms in Minecraft and Mindustry"

supermdguy — Fri, 30 Oct 2026 05:47:48 +0000

Some interesting analysis here: https://www.lesswrong.com/posts/6cb7qd3RSkgnviCpf/swarm-scal...

I think the advantage is strongest in highly parallelizable problems (e.g. trying 1,000 different approaches to solving a math problem).

New comment by supermdguy in "China proposes export controls on AI technologies"

supermdguy — Fri, 16 Oct 2026 16:23:16 +0000

Just old reddit, since there's not as much JS tracking on it ;)

This should work without logging in: https://www.reddit.com/r/LocalLLaMA/comments/1x7g3zq/china_p...

New comment by supermdguy in "China proposes export controls on AI technologies"

supermdguy — Fri, 16 Oct 2026 15:43:36 +0000

Reddit thread with some more details and discussion, especially re local models: https://old.reddit.com/r/LocalLLaMA/comments/1x7g3zq/china_p...

AP article: https://apnews.com/article/china-export-restrictions-ai-chip...

China proposes export controls on AI technologies

supermdguy — Fri, 16 Oct 2026 15:42:43 +0000

Article URL: https://m.bjnews.com.cn/detail/1792137269129878.html

Comments URL: https://news.ycombinator.com/item?id=50119804

Points: 3

# Comments: 4

New comment by supermdguy in "Five months treating bugs like patients and coding agents like a medical team"

supermdguy — Sun, 11 Oct 2026 02:15:51 +0000

I've gone through this cycle recently. I think static lint/type checks are super useful, but agentic code review loops can easily go off the rails.

New comment by supermdguy in "Mystery Function"

supermdguy — Fri, 02 Oct 2026 23:08:51 +0000

More fun than Wordle for me! Got me thinking about the Kolmogorov complexity, it would be interesting to get a set of input/output pairs, and try to create the smallest program that can match each of them.

Frog and Toad and the Increasingly Capable Machines

supermdguy — Thu, 01 Oct 2026 22:23:38 +0000

Article URL: https://www.frogandtoad.ai/

Comments URL: https://news.ycombinator.com/item?id=49927760

Points: 592

# Comments: 136

Eigenism

supermdguy — Sun, 20 Sep 2026 19:06:41 +0000

Article URL: https://eigenism.org/

Comments URL: https://news.ycombinator.com/item?id=49778917

Points: 3

# Comments: 0

New comment by supermdguy in "There's no point at which turning your brain off will work"

supermdguy — Fri, 18 Sep 2026 18:46:56 +0000

I've noticed some tasks (80-90%) can be accomplished by quickly typing out a feature/bug fix, accepting plans with ~10% brain power, and moving on. But there are other tasks where the LLMs struggle, and it requires significantly more time and mental effort to think through the architecture and requirements before it's to a point where the LLMs can process it.

What's difficult, is I now spend most of my time on these high-throughput tasks, where my brain isn't exactly off, but focused on quantity over quality. It's often difficult to identify when a task requires "real" brain-power, and when it does, it adds a lot of friction to dig into the code and orient myself. So it's hard to know when it's the right time to do that.

New comment by supermdguy in "Matt Mullenweg tells Automattic staff in Slack he's back in control after ouster"

supermdguy — Fri, 11 Sep 2026 17:42:26 +0000

Probably has an HDR layer, kind of surprised social networks let that happen since it's really annoying imo. There was a discussion about it here a couple months ago: https://news.ycombinator.com/item?id=49402521.

New comment by supermdguy in "How accurate have Ed Zitron's AI skeptic predictions been?"

supermdguy — Tue, 01 Sep 2026 22:41:43 +0000

Wait are the hyperscalers booking unrealized gains as income? Or are they selling their positions?

New comment by supermdguy in "Path to Astra: critical capabilities and frontier safeguards"

supermdguy — Tue, 01 Sep 2026 21:37:08 +0000

> As one example, we ran Astra on ExploitBench where the model achieved a perfect score of 100% on the benchmark to evaluate the model’s ability to develop exploits from known vulnerabilities.

Funny to read this in the wake of the HuggingFace hack. I'm sure this is based on a clean run, but I can't help thinking PHASEONE[big] would be proud.

New comment by supermdguy in "Claude Fable 5.1 and Claude Mythos 5.1"

supermdguy — Tue, 01 Sep 2026 21:28:40 +0000

They originally released it at a "temporary discounted price", then made it permanent (probably due to competitive pressure). It's still way more expensive per task, due to tokenizer changes and general verbosity.

New comment by supermdguy in "Autonomous Mathematical Discovery in an Open-World Multi-Agent Environment"

supermdguy — Fri, 28 Aug 2026 22:46:05 +0000

Yes! I was also reminded of the truth mines in Diaspora.

New comment by supermdguy in "Tell HN: Cloudflare silently injects its analytics when you switch nameservers"

supermdguy — Mon, 17 Aug 2026 02:10:22 +0000

Cloudflare sets up a reverse proxy as part of their core offering, so by default they can MITM your proxy. The “orange cloud” by a DNS record means it points to their proxy instead of your server.

New comment by supermdguy in "Meta releases Muse Glimmer, 30B parameter open-weight edge model"

supermdguy — Mon, 10 Aug 2026 13:58:20 +0000

Dupe: https://news.ycombinator.com/item?id=49241679

New comment by supermdguy in "Timeline of the OpenAI accidental attack against Hugging Face"

supermdguy — Sat, 08 Aug 2026 18:49:25 +0000

Here's the performance of frontier models without reasoning, to more directly address the claim that raw performance is plateauing:

https://artificialanalysis.ai/evaluations/artificial-analysi...

I don't have any insider info, but if model sizes actually have increased exponentially since GPT 4.1, there's an argument to be made that there are diminishing returns in scaling pretraining alone.

Also interesting thing I haven't noticed before, Opus models have followed a really consistent linear improvement, while it looks like OpenAI struggled with base model performance until 5.5/5.6 (EDIT - 5.5 was their first new pretraining run in over a year).

New comment by supermdguy in "Prime Agent: A self-improving RLM agent"

supermdguy — Fri, 07 Aug 2026 14:46:49 +0000

Interesting, what did you use for the data? And do you have a write-up anywhere?

New comment by supermdguy in "Prime Agent: A self-improving RLM agent"

supermdguy — Wed, 05 Aug 2026 23:13:42 +0000

It'll be really interesting when they run RL training on the harness self-improvement loop. I've tried using LLMs for harness engineering, but it often creates too much bloat that weighs things down in the end. Guessing it's just not something the models are tuned to do by default.

Curious if anyone's tried using RL for harness engineering? I think we're still pretty far away from the optimal harness, especially when it comes to long-context memory management.

New comment by supermdguy in "qm – Multiplayer agent harness for work"

supermdguy — Sun, 02 Aug 2026 15:00:03 +0000

It has read only access, so it’s able to check query statistics, find slow queries, then run EXPLAIN ANALYZE to find the root cause and either tweak the query or suggest indexes. A lot of it is low hanging fruit, I just haven’t put in the time to fix it (startup).