Hacker News: petra

New comment by petra in "Discovery Loop"

petra — Thu, 06 Aug 2026 01:15:32 +0000

Creation of nuclear reactors that are useless for terrorists could help

New comment by petra in "Discovery Loop"

petra — Thu, 06 Aug 2026 01:12:18 +0000

Software engineers are in a decades long project of automating everyone and everything else, and getting the financial rewards out of that.

So...

New comment by petra in "LLMs reward expertise"

petra — Tue, 04 Aug 2026 04:55:03 +0000

So maybe your friend need a role reversal: A prompt guiding the llm to act as a consultant, guiding people in implementing a software project, asking questions, creating a shared understanding, limiting scope or creating milestones.

That seems like something that could be done using an llm, not that complicated probably.

And maybe, in other fields as well.

New comment by petra in "Discovering Cryptographic Weaknesses with Claude"

petra — Wed, 29 Jul 2026 06:34:56 +0000

They probably took our skills and reuse them. That's why skills do not matter anymore.

New comment by petra in "Discovering Cryptographic Weaknesses with Claude"

petra — Wed, 29 Jul 2026 06:32:43 +0000

It can follow instructions well. Just give it a spec in a formal, well defined language. runnable language and some reinforcement learning.

It can also handle vague stuff.

That's certainly more powerful than a regular computer language.

New comment by petra in "EYG: A Programming Language for Humans"

petra — Tue, 28 Jul 2026 05:00:48 +0000

// makers want to make not ask ai to make,

It probably changes from person to person, some people are just interested in seeing how their idea comes to life and maybe does something useful in the world.

New comment by petra in "Microsoft is using TPM chips to crack down on pirated Windows activations"

petra — Mon, 27 Jul 2026 17:31:50 +0000

Also weird old hardware.

New comment by petra in "Show HN: Descript wanted $24/mo, I built an open-source alternative in a weekend"

petra — Mon, 27 Jul 2026 15:59:27 +0000

But this guy will probably get hired because of this.

New comment by petra in "Chinese chipmaker shares surge 470%"

petra — Mon, 27 Jul 2026 15:01:15 +0000

And besides llm's memory requirements:

-HBM has a lot of room to grow(bandwidth/power/density). And an HBM memory takes 3x the amount of wafers of regular DRAM.

-HBM + better DRAM designs could increase random memory access speed by alot. I've seen research talking about 7x.

At what point it makes sense that half of consumer GPUs and most of the server GPU/CPUs start using HBM?

New comment by petra in "The New AI Superpowers: Focus and Followthrough"

petra — Mon, 27 Jul 2026 10:07:37 +0000

Given the limits of LLM's regarding context, why is zero dependencies the goal ?

Shouldn't something like maximum reuse/abstraction, minimal written code, should be a better goal ?

Possible leading to more llm-scalable systems ?

New comment by petra in "AMD's 256-core Epyc 9996 'Venice' claims up to a 3.4x jump over Intel Xeon"

petra — Thu, 23 Jul 2026 21:05:31 +0000

The other side of llm's is the need for rich feedback, often by simulations of various things. And memory latency is a big bottleneck there.

So this chip nicely complements the llm's.

New comment by petra in "Does creatine make you smarter?"

petra — Wed, 22 Jul 2026 16:31:46 +0000

LoL.maybe the question is not phrased in the best way, English as a second language and laziness, but it's a pretty decent question.

New comment by petra in "Does creatine make you smarter?"

petra — Wed, 22 Jul 2026 15:54:13 +0000

Given LLM's is being smarter still a worthwhile to optimize with supplements?

New comment by petra in "AMD Ryzen 7 7700X3D Review: 3D V-Cache Gaming Performance for Less – HotHardware"

petra — Wed, 22 Jul 2026 12:01:49 +0000

I wonder what perf. this cpu can get in bit/ternary inference, and whether the 512AVX units(or whats it's called on AMD), get throttled.

New comment by petra in "Ten Steps Towards Happiness (2015)"

petra — Wed, 22 Jul 2026 06:48:07 +0000

You're right. I've used reader mode(the paper in the url bar in Firefox) to solve this.

New comment by petra in "Israel is building a miles-long barrier in Gaza, entrenching its division"

petra — Wed, 22 Jul 2026 05:46:19 +0000

That's the current Netanyahu and right wing extremists government. Nothing good can be expected of them.

Elections are going to be held soon. Hopefully this government will be replaced, and better things start happening.

New comment by petra in "Biggest Probabilistic Computer Turns Noise into Answers"

petra — Mon, 20 Jul 2026 03:37:29 +0000

There's also digital probabilistic computing, with a processor working on AWS, and for the right workload ,claimed to be offering a very significant speed boost:

https://www.eetimes.com/probabilistic-computing-is-already-h...

Towards the end of the article.

New comment by petra in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

petra — Sat, 18 Jul 2026 20:48:42 +0000

We already drown in useful new things and especially knowledge. There's a limit to how fast we can adopt them.

New comment by petra in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

petra — Sat, 18 Jul 2026 20:42:46 +0000

Llm's do have discipline and persistence, though.

New comment by petra in "GPT-5.6 used a prompt to close a 30-year gap in convex optimization"

petra — Sat, 18 Jul 2026 20:33:07 +0000

The ability to correct the plan/prompt.