Hacker News: Schlagbohrer

New comment by Schlagbohrer in "Golden Testing a CAD Library"

Schlagbohrer — Thu, 14 May 2026 07:06:45 +0000

When it comes to LLM's doing adversarial training, I wonder if there is any value to letting them rapidly design and simulate circuits or mechanical models.

New comment by Schlagbohrer in "Princeton mandates proctoring for in-person exams, upending 133 year precedent"

Schlagbohrer — Thu, 14 May 2026 07:04:29 +0000

Why did the other comment by s5300 get downvoted literally to death? Many people who have suffered first hand at the depravity of the american medical system are right to be furious about it. Is this forum full of cardiologists and medical staff?

New comment by Schlagbohrer in "US producer prices surprise with largest increase in four years"

Schlagbohrer — Wed, 13 May 2026 19:54:12 +0000

Yeah quite a surprise, between the war on Iran plus unregulated profiteering.

New comment by Schlagbohrer in "What's a mathematician to do? (2010)"

Schlagbohrer — Sun, 10 May 2026 13:46:12 +0000

After reading another post about the most recent advances LLMs have made in finding and writing up novel, correct proofs, it sounds like the frontier models are now at the point of PhD student level. I wonder how a math student could contribute today, if they're just starting on the PhD track? Maybe by using LLMs as a mighty tool and providing skilled usage and oversight?

It must feel similar to those who wanted to become chess or go masters after computers surpassed humanity in those games.

New comment by Schlagbohrer in "Casio S100X Japanese Lacquer Edition (JP Page Only)"

Schlagbohrer — Sun, 10 May 2026 13:43:36 +0000

Absolutely have zero use for it, 100% feel like i MUST have one

Comparing a 1980s memory map to the Raspi Pico

Schlagbohrer — Sun, 10 May 2026 12:58:11 +0000

Article URL: https://medium.com/@noborutakahashi/a-40-year-old-memory-map-comparable-to-todays-raspberry-pi-pico-932c4309260d

Comments URL: https://news.ycombinator.com/item?id=48083622

Points: 27

# Comments: 0

New comment by Schlagbohrer in "An Introduction to Meshtastic"

Schlagbohrer — Sat, 09 May 2026 19:23:29 +0000

Useful for things like music festivals, campgrounds, having a mesh node in your camper van and another with your friends and another on your person. Imagine a large festival outside of good cellphone coverage or a camp site in the middle of nowhere. There are also credit-card sized versions which can be put in a dog's harness in case the dog wanders off (the radio contains a GPS).

New comment by Schlagbohrer in "Maybe you shouldn't install new software for a bit"

Schlagbohrer — Fri, 08 May 2026 20:37:16 +0000

I juuuuuust updated npm, should i be worried?

New comment by Schlagbohrer in "Valve releases Steam Controller CAD files under Creative Commons license"

Schlagbohrer — Thu, 07 May 2026 14:14:21 +0000

As long as we are doing controller chat, I want to reminisce about the original Xbox Elite controller with replaceable batteries. Very expensive but the weight was nice, the replaceable buttons and joysticks were great, and the replaceable batteries meant that a rechargable battery wasn't going to degrade overtime and make it unusable.

Unfortunately it had some real flaws. The back buttons weren't supported even on windows, I had to use 3rd party software like REWASD to map them to keyboard keys. And despite the high cost of the device, the rubber on it degraded and broke off pretty quickly. And the controller didn't have backlit buttons!

New comment by Schlagbohrer in "Valve releases Steam Controller CAD files under Creative Commons license"

Schlagbohrer — Thu, 07 May 2026 14:11:32 +0000

New comment by Schlagbohrer in "Google Chrome silently installs a 4 GB AI model on your device without consent"

Schlagbohrer — Tue, 05 May 2026 10:56:52 +0000

Old heads checking in... Back in my day, we had an exposed file hierarchy and we liked it!

New comment by Schlagbohrer in "The Road to a Billion-Token Context"

Schlagbohrer — Mon, 04 May 2026 10:34:23 +0000

Amazing that they are trying to solve this with hardware rather than with a new software architecture but I suppose the current technology underlying LLM software must be far and away the best theoretically or most established, or the time taken to seek a new model isn't worth it for the big companies.

I know Yann LeCun is trying to do a completely different architecture and I think that's expected to take 2-3 years before showing commercial results, right? Is that why they're finding it quicker to change the hardware?

New comment by Schlagbohrer in "The Road to a Billion-Token Context"

Schlagbohrer — Mon, 04 May 2026 10:32:49 +0000

What does this mean: "In addition, because most AI models are not trained uniformly across their maximum context length, their reasoning quality tends to degrade gradually near the limit rather than fail abruptly."

Models aren't trained across their context, their context is their short term memory at runtime, right? Nothing to do with training. They are trained on a static dataset.

New comment by Schlagbohrer in "Ti-84 Evo"

Schlagbohrer — Sat, 02 May 2026 15:20:42 +0000

I always thought they should make a thin, metal, foldable 83 variant that just bends in the middle and looks like a cigarette case.

New comment by Schlagbohrer in "David Silver of DeepMind raises $1B to build AI that learns without human data"

Schlagbohrer — Tue, 28 Apr 2026 08:48:51 +0000

Geoffrey Hinton said that the breakthrough the AlphaGo team had was getting it to play against itself and improve in that means, since it could then go beyond the human training data it had learned on. He said that an equivalent form of self-training for generalized information would let a superintelligence take off (this is from my memory, not an exact quote).

The TechCrunch article doesn't specify how/what kind of data a recursive general AI could use to achieve such a thing. If it is possible that's exciting. Seems like a real philosophical question to answer- How could a general AI self-train?

New comment by Schlagbohrer in "David Silver of DeepMind raises $1B to build AI that learns without human data"

Schlagbohrer — Tue, 28 Apr 2026 08:46:34 +0000

"pre-money valuation" I don't know what that means but it makes me roll my eyes so hard it hurts

New comment by Schlagbohrer in "4TB of voice samples just stolen from 40k AI contractors at Mercor"

Schlagbohrer — Mon, 27 Apr 2026 14:38:18 +0000

Tell us more about that fraud story! Was the person your attorney or accountant? Or just some "smart" person who decided to wisely save time by doing fraud?

New comment by Schlagbohrer in "Toward an Ethics and Etiquette for Electronic Mail (1985)"

Schlagbohrer — Mon, 27 Apr 2026 14:30:45 +0000

Excerpt:

The Phenomenon of "Flaming"

Perhaps the attribute of electronic mail systems that most distinguishes them from other forms of communication is their propensity to evoke emotion in the recipient — very likely because of misinterpretation of some portion of the form or content of the message — and the likelihood that the recipient will then fire off a response that exacerbates the situation.

Toward an Ethics and Etiquette for Electronic Mail (1985)

Schlagbohrer — Mon, 27 Apr 2026 14:30:45 +0000

Article URL: https://www.rand.org/pubs/reports/R3283.html

Comments URL: https://news.ycombinator.com/item?id=47922153

Points: 2

# Comments: 1

New comment by Schlagbohrer in "Microsoft and OpenAI end their exclusive and revenue-sharing deal"

Schlagbohrer — Mon, 27 Apr 2026 14:25:49 +0000

Especially when the OpenAI definition of AGI is only in financial terms (when it becomes profitable), which can be easily manipulated.