Hacker News: bethekind

New comment by bethekind in "Sawe becomes first athlete to run a sub-two-hour marathon in a competitive race"

bethekind — Mon, 27 Apr 2026 00:14:28 +0000

Where does discussion on gut training occur? All I know is you need a 5:4 ratio of glucose to fructose? Then when you train, you use the gels and the more you do it, the more capable your gut gets at absorbing without distress.

Is that all the science to it?

New comment by bethekind in "The electromechanical angle computer inside the B-52 bomber's star tracker"

bethekind — Sun, 19 Apr 2026 00:06:25 +0000

Off topic, but this is where I see AI going. A tool that condenses work down from requiring a team and a room to a box. We're decades away from that

New comment by bethekind in "Generative art over the years"

bethekind — Fri, 10 Apr 2026 06:09:29 +0000

My favorite is Brownian motion fractal art. Something about it just tickles my brain just right

New comment by bethekind in "Netflix Prices Went Up Again – I Bought a DVD Player Instead"

bethekind — Thu, 09 Apr 2026 21:58:01 +0000

Me too

New comment by bethekind in "Claude Code is unusable for complex engineering tasks with the Feb updates"

bethekind — Mon, 06 Apr 2026 19:47:08 +0000

1 client, 1 agent? Interesting

New comment by bethekind in "Codex pricing to align with API token usage, instead of per-message"

bethekind — Mon, 06 Apr 2026 02:42:31 +0000

Poolometer looks cool! I will say your smiley face icons look a lil odd and ai generated, but otherwise I love the graph tracking and suggestions

New comment by bethekind in "Codex pricing to align with API token usage, instead of per-message"

bethekind — Sun, 05 Apr 2026 19:12:06 +0000

I agree. Gemini models are held back by their segmentation of usage between multiple products, combined with their awful harnesses and tooling. Gemini cli, antigravity, Gemini code assist, Jules.... The list goes on. Each of these products has only a small limit and they must share usage.

It gets worse than that though. Most harnesses that are made to handle codex and Claude cannot handle Gemini 3.1 correctly. Google has trained Gemini 3.1 to return different json keys than most harnesses expect resulting in awful results and failure. (Based on me perusing multiple harness GitHub issues after Gemini 3.1 came out)

New comment by bethekind in "Emotion concepts and their function in a large language model"

bethekind — Sat, 04 Apr 2026 07:12:25 +0000

They likely already have. You can use all caps and yell at Claude and it'll react normally, while doing do so with chatgpt scares it, resulting in timid answers

New comment by bethekind in "Tell HN: Anthropic no longer allowing Claude Code subscriptions to use OpenClaw"

bethekind — Sat, 04 Apr 2026 00:52:22 +0000

I think my next steps are: 1) try out openai $20/month. I've heard they're much more generous. 2) try out open router free models. I don't need geniuses, so long as I can see the thinking (something that Claude code obfuscates by default) I should be good. I've heard good things about the CLIO harness and want to try openrouter+clio

New comment by bethekind in "Claude Code's source code has been leaked via a map file in their NPM registry"

bethekind — Tue, 31 Mar 2026 14:41:13 +0000

This. If I run 4 Claude code opus agents with subagents, my 8gb of RAM just dies.

I know they can do better

New comment by bethekind in "$500 GPU outperforms Claude Sonnet on coding benchmarks"

bethekind — Fri, 27 Mar 2026 14:29:03 +0000

I've used Gemini and now claude. Both were meh until I found the superpowers skill. Will be trying chatgpt next month.

You can "feel" the llm being limited with Gemini, less so with Claude. Hopefully even less so with chatgpt

New comment by bethekind in "I tried Karpathy's Autoresearch on an old research project"

bethekind — Mon, 23 Mar 2026 19:41:16 +0000

I used it to speed up an codecompass-like repo from 86 files per second to 2000. Still haven't used the repo in production, so maybe it secretly broke things, but the ability to say: "optimize this benchmark and commit only if you pass these tests" is nice

New comment by bethekind in "Launch HN: IonRouter (YC W26) – High-throughput, low-cost inference"

bethekind — Fri, 13 Mar 2026 00:05:00 +0000

I love the phrase "inference hunger"

New comment by bethekind in "A case for Go as the best language for AI agents"

bethekind — Mon, 02 Mar 2026 21:35:20 +0000

Jane Street had a cool video about how you can address lack of training data in a programming language using llm patching. Video is called "Arjun Guha: How Language Models Model Programming Languages & How Programmers Model Language Models"

The big take away is that you can "patch" llms and steer them to correct answers in less trained programming languages, allowing for superior performance. Might work here. Not a clue how to implement, but stuff to llm-to-doc and the like makes me hopeful

New comment by bethekind in "Can you reverse engineer our neural network?"

bethekind — Fri, 27 Feb 2026 13:49:18 +0000

Model interpretability is going to be the final frontier of software. You used to need to debug the code. Now you'll need to debug the AI.

New comment by bethekind in "Google restricting Google AI Pro/Ultra subscribers for using OpenClaw"

bethekind — Sun, 22 Feb 2026 23:58:46 +0000

This is draconian.

> Our investigation specifically confirmed that the use of your credentials within the third-party tool “open claw” for testing purposes constitutes a violation of the Google Terms of Service [1]. This is due to the use of Antigravity servers to power a non-Antigravity product. I must be transparent and inform you that, in accordance with Google’s policy, this situation falls under a zero tolerance policy, and we are unable to reverse the suspension. I am truly sorry to share this difficult news with you.

New comment by bethekind in "DeepSeek OCR"

bethekind — Mon, 20 Oct 2025 07:13:39 +0000

Did we read the same graph? DeepSeek Gundam 200 dpi appeared to get similar perf as dots-ocr, but with less tokens needed. The x axis is inverted, descending with distance from the origin.

New comment by bethekind in "The Math of Catastrophe"

bethekind — Fri, 19 Sep 2025 04:49:44 +0000

> the behavior of a complex system can simplify as it passes from one state to another. “Sometimes a high-dimensional system can tip,” said Lenton, “and when it gets near tipping, it starts to behave like a much lower-dimensional system.” The lesson, he added, echoes the one learned at Peter Lake: to “simplify without oversimplifying.”

Sounds a lot like what people WANT neural networks to do. Collapse a high dimensional situation into a very low dimensional network. Ideally a binary answer, yes or no. I wonder if this bifurcation chaos math has implications in ml work

New comment by bethekind in "Ozempic shows anti-aging effects in trial"

bethekind — Wed, 06 Aug 2025 06:05:35 +0000

If Eli Lily is the only producer, how is the gray market being supplied? This makes no sense...

That being said, I'm waiting for oral GLP1 agonists. Injections are a hassle and gray market ones even more so

New comment by bethekind in "Julia Parsons, U.S. Navy Code Breaker During World War II, Dies at 104"

bethekind — Thu, 01 May 2025 07:17:00 +0000

> ...as much as it is for anyone....

The phrasing in this sentence implies that the Julia language could be named for the code breaker, as much as it could be named for anyone else. In other words, it wasn't named for the code breaker, but it might as well have been.

The follow up comment gives hard quantitative fact that the language wasn't named for anyone or anything. I can see how both comments are correct, the first implicitly, the second explicitly