Hacker News: 1R053

New comment by 1R053 in "We put a coding agent in a while loop"

1R053 — Sun, 24 Aug 2025 20:02:32 +0000

I guess pkill would rather be a sleep or koma. Erasing itself from any storage would rather equate to aicide

New comment by 1R053 in "Exit Tax: Leave Germany before your business gets big"

1R053 — Thu, 07 Aug 2025 19:47:12 +0000

While that number seems to be not a general value, the "Wegzugsbesteuerung" still is significant.

https://de.wikipedia.org/wiki/Wegzugsbesteuerung

Essentially, it assumes you sell your assets at market value and taxes the difference to your expenses for it.

New comment by 1R053 in "Show HN: I built a synthesizer based on 3D physics"

1R053 — Sat, 03 May 2025 09:20:46 +0000

I think to be useful it needs a mode of playing, that is always musical / in tune.

Not yet sure how to really do it, but one concept I like from NI plugins is that you have multiple keyboard zones: one zone is for notes, others are e.g. for patterns or styles. Imagine a guitar where one zone is for the chord type and tone, another for the striking pattern...

The challenge here is probably the resonance algo for multiple systems based on multiple notes... Maybe the piano concept would be handy here... imagine instead of having 3 strings like on the piano the instrument to be one system for each key... that excite each other via air or direct resonance points... the systems should be automatically tuned based on one reference system (e.g. using automatic string length or tension scaling)

Anyway, amazing work and having it on GPU allows this really to scale.

New comment by 1R053 in "Apple M3 Ultra"

1R053 — Thu, 06 Mar 2025 09:39:21 +0000

are you comparing the same models? How did you calculate the TOPS for M3 Ultra?

New comment by 1R053 in "AI is killing some companies, yet others are thriving – let's look at the data"

1R053 — Sat, 01 Mar 2025 06:57:35 +0000

I think it depends a lot on how research AI generated traffic counts towards these stats.

Also, if there is no answer yet on the web the AI may also not know it. Then these questions should still end up on SO.

I might add, that SO also could build their own chat / research UI. It would need to have some benefit over others, but I guess the community aspect of it alone would suffice...

New comment by 1R053 in "LIMO: Less Is More for Reasoning"

1R053 — Mon, 10 Feb 2025 11:11:51 +0000

I think the title of the paper is misleading. Obviously the result shows an impressive performance with just few training examples. However, I cannot see that while keeping the same method reducing training data leads to more performance. They have simply shifted the performance curve (impressively) to lower thresholds. Still also with this new method more training data should give better results. It would be interesting to see a full performance curve for the method based on training data amount (and potentially quality).

New comment by 1R053 in "OpenAI: Out of Ideas"

1R053 — Mon, 03 Feb 2025 19:36:31 +0000

depending on whether the incumbents will keep the ability to ride new S-curves...

as in https://en.wikipedia.org/wiki/The_Innovator%27s_Dilemma

New comment by 1R053 in "OpenAI: Out of Ideas"

1R053 — Mon, 03 Feb 2025 18:24:58 +0000

not sure if there would be a deepseek open source model if there were no leading models from OpenAI et al

New comment by 1R053 in "Science YouTuber physicsgirl (Dianna Cowern) stands for the first time in 2 yrs"

1R053 — Wed, 29 Jan 2025 11:24:52 +0000

that is a horrible way to be treated by a doctor. In Germany you have free choice of your doctors. Although sometimes you have to wait quite a bit to get appointments... I am not sure, but with a a European health insurance card you should be able to also go to doctors in other countries and be at least partially covered by your insurance.

New comment by 1R053 in "Mathics 7.0 – Open-source alternative to Mathematica"

1R053 — Sat, 07 Dec 2024 15:36:28 +0000

one of the annoying things about Mathematica is that all functions are crammed into the same namespace and that there is no overloading with different parameterization options...

New comment by 1R053 in "Elon Musk and Vivek Ramaswamy: The Doge Plan to Reform Government"

1R053 — Thu, 21 Nov 2024 09:54:29 +0000

creative destruction is not necessarily bad, but the default way of nature to move forward.

Obviously, you need to destroy the right things...

New comment by 1R053 in "Elon Musk and Vivek Ramaswamy: The Doge Plan to Reform Government"

1R053 — Thu, 21 Nov 2024 09:52:32 +0000

To everyone voting it was clear, that Elon would get this role. So he was part of the elected package. Probably, that was also a deciding factor for many voters.

New comment by 1R053 in "Tencent Hunyuan-Large"

1R053 — Wed, 06 Nov 2024 08:39:49 +0000

An H100 supports 80 GB of memory. so at FP8 that would allow 3 of the 16+1 models per GPU (assuming around 26B per model), requiring 9 H100s, that usually would not fit one chassis I guess.

Once you have something with 192 GB it gets interesting. You could probably have 7 at FP8 per GPU. At FP16 it probably only would fit 3 per card, requiring 9 again.

I'd say for the current memory layout of cards they missed a little bit the sweet spot. With slightly smaller models or one expert less one should be able to run it on 8 H100s at FP8 or 2 B100s at FP8 or even on 4 B100s at FP16 if I calculated correctly.

New comment by 1R053 in "Tencent Hunyuan-Large"

1R053 — Tue, 05 Nov 2024 20:37:45 +0000

you can get a long way on something with 41% less performance than your favorite supercar...

New comment by 1R053 in "Tencent Hunyuan-Large"

1R053 — Tue, 05 Nov 2024 20:31:36 +0000

the paper with details: https://arxiv.org/pdf/2411.02265

They use

- 16 experts, of which one is activated per token

- 1 shared expert that is always active

in summary that makes around 52B active parameters per token instead of the 405B of LLama3.1.

New comment by 1R053 in "Former Bayesian captain on sinking, vessel ballast and stability"

1R053 — Tue, 05 Nov 2024 10:09:00 +0000

how fast do these machineries move? as I understand the weather scenario the wind direction was changing rapidly by more than 90 degrees at hurricane level. That means, the boat was unable even under power to redirect into the wind timely enough. The key question is, how much side wind can the design withstand without getting pressed into more than 45 degrees.

obviously lowering the keel affects this a lot, as there is much difference in leveraged 60 tons vs non-leveraged.

New comment by 1R053 in "M4 MacBook Pro"

1R053 — Wed, 30 Oct 2024 15:53:07 +0000

probably the next update wave is coming from the need of AI features for more local memory and compute. The software is just not there yet in usual tasks but it's just a question of time I guess. Of course there will be the pressure to do that in the cloud as usual, but local compute will always remain a market.

and probably it's good that at least one of the big players has a business model that supports driving that forward

New comment by 1R053 in "The Remarkable Life of Ibelin"

1R053 — Sat, 05 Oct 2024 18:17:19 +0000

I do empathise with your desire a lot

...but then such a service gets monetised by advertising, financed with VC money and guess what happens next...

New comment by 1R053 in "Show HN: I made an open-source Notion-style WYSYWIG editor"

1R053 — Fri, 16 Jun 2023 23:44:18 +0000

Can you recommend a good tutorial?

New comment by 1R053 in "Ask HN: What kind of electronics would you stockpile for doomsday?"

1R053 — Tue, 16 Nov 2021 00:07:02 +0000

The PV on the roof probably wont help anymore after am EMP. So you‘d need the whole package incl water purifier also in a shielded box