Hacker News: sojuz151

New comment by sojuz151 in "Kraków, Poland in top 5 worst air quality worldwide"

sojuz151 — Tue, 20 Jan 2026 08:36:53 +0000

The air in Kraków is fine once you give it a good chew. I don't know why people are complaining.

New comment by sojuz151 in "Solving the Problems of HBM-on-Logic"

sojuz151 — Tue, 23 Dec 2025 17:47:03 +0000

I don't understand why even put that HBM on top of the core.

From what I understand, in a typical gpu core you put logic and connectors on one side and innert silicon on the other. So unless you drill through silicon you don't get shorter routing.

Why not put GPU one one side and HBM on the other side of the PCB? This would fix the cooling problem?

New comment by sojuz151 in "TPUs vs. GPUs and why Google is positioned to win AI race in the long term"

sojuz151 — Thu, 27 Nov 2025 15:35:45 +0000

They lose the competitive advantage. They have nothing more to offer than what Google has in-house.

New comment by sojuz151 in "Should LLMs just treat text content as an image?"

sojuz151 — Mon, 27 Oct 2025 13:43:34 +0000

This means that current tokenisers are bad, and something better is needed if text rendering + image input is a better tokeniser.

New comment by sojuz151 in "Quantitative AI progress needs accurate and transparent evaluation"

sojuz151 — Fri, 25 Jul 2025 12:15:29 +0000

Compute has been getting cheaper and models more optimised. So if models can do something it will not be long till they can do this cheap.

New comment by sojuz151 in "Eleven v3"

sojuz151 — Thu, 05 Jun 2025 20:03:10 +0000

Polish is quite good, expected based on the founders' background

New comment by sojuz151 in "Can rotation solve the Hubble Puzzle?"

sojuz151 — Wed, 23 Apr 2025 15:17:30 +0000

And the Hubble tension is around 4%, within a factor of 2 of that 10%. This sounds reasonable.

New comment by sojuz151 in "Carbon capture more costly than switching to renewables, researchers find"

sojuz151 — Sun, 16 Feb 2025 00:03:26 +0000

Also, with renewable energy sources, the problem is not manufacturing energy per se but getting the energy where and when you want.

For example, you could run a gas turbine in Germany in winter at night during a spike of the consumption and run carbon capture on a desert during the day

In a sense you use atmosphere as battery/transportation system.

I am not saying that this must make sense but this is far more complex than napkin math.

New comment by sojuz151 in "Why the weak nuclear force is short range"

sojuz151 — Wed, 15 Jan 2025 18:52:17 +0000

Because in a classical theory, where there are no particles, there is still the same short range potential.

New comment by sojuz151 in "How does it feel to test a compiler?"

sojuz151 — Fri, 16 Aug 2024 08:22:40 +0000

What about using multiple open source projects with already existing tests? You check if results of execution are the same between compiler versions? This should also provide you with a better coverage of various features

New comment by sojuz151 in "Ask HN: How do browsers isolate internal audio from microphone input?"

sojuz151 — Tue, 09 Jul 2024 21:41:09 +0000

Remember that convolution is multiplication in the frequency domain, so this also handles different responses at different frequencies, not just delays

New comment by sojuz151 in "TDK claims solid state battery breakthrough"

sojuz151 — Mon, 17 Jun 2024 17:57:59 +0000

This will have same energy density as TNT. I have a feelings that fully chargred battery of this type might explode

New comment by sojuz151 in "Gemini Flash"

sojuz151 — Wed, 15 May 2024 11:47:46 +0000

>My intuition is that as contexts get longer we start hitting the limits of how much comprehension can be embedded in a single point of vector space, and will need better architectures for selecting the relevant portions of the context.

We are dealing with multi-headed attention, therefore we have multiple points per token. You can always increase the number of heads or the size of the key vector.

New comment by sojuz151 in "Ointers: A library for representing pointers where bits have been stolen (2021)"

sojuz151 — Wed, 08 May 2024 12:05:48 +0000

This look like something that debuggers would hate, because following pointers would be broken

New comment by sojuz151 in "LPCAMM2 is a modular, repairable, upgradeable memory standard for laptops"

sojuz151 — Tue, 07 May 2024 16:37:24 +0000

I would say it would make the most sense to have a replaceable entire ram+cpu+gpu assemble. Just have some standard form factors and connectors for external connectors.

This way, you could keep power consumption low and be able to upgrade cpu to a new generation

New comment by sojuz151 in "Scientists discover first nitrogen fixing organelle"

sojuz151 — Sun, 21 Apr 2024 14:31:24 +0000

I don't think so. Biological nitrogen fixation is far too inefficient for this to make sense. Deceased in yield would be too big tfor this to make sense

To give some numbers, just the fixation takes 5.4 more energy, and photosynthesis is 6 times less efficient than solar panels. Moving to biological fixation would also increase water and pesticide usage

New comment by sojuz151 in "Show HN: Not sure you're talking to a human? Create a human check"

sojuz151 — Tue, 19 Mar 2024 15:55:19 +0000

Also, you can use this to supplement income. You can herd goat and solve capcha at the same time.

New comment by sojuz151 in "The end of the dark universe?"

sojuz151 — Fri, 15 Mar 2024 23:47:34 +0000

This theory is trying to justify MOND, and MOND by itself has many problems. It has problems with explaining the bullet cluster, CMB spectrum, and rotation curves of some galaxies. This paper was not even peer reviewed. This is almost certainly nothing burger

New comment by sojuz151 in "Apple terminates Epic Games developer account, calling it a 'threat' to iOS"

sojuz151 — Wed, 06 Mar 2024 18:54:33 +0000

The lose of revenue would not be the biggest problem, they could afford it. The problem would be the loss of applications, most EU based developers would stop supporting iOS and focus on Android. That would hurt

New comment by sojuz151 in "A quantitative test of Diamond’s axis of orientation hypothesis"

sojuz151 — Thu, 15 Feb 2024 15:33:54 +0000

My biggest problem is that this hypothesis does not explain why farming started so much later in the Americas than in Eurasia. From Wikipedia "approximate centres of origin of agriculture and its spread in prehistory: the Fertile Crescent (11,000 BP), the Yangtze and Yellow River basins (9,000 BP) and the Papua New Guinea Highlands (9,000–6,000 BP), Central Mexico (5,000–4,000 BP), Northern South America (5,000–4,000 BP), sub-Saharan Africa (5,000–4,000 BP, exact location unknown), eastern North America (4,000–3,000 BP). " and American civilisation were about 4 thousand years behind Eurasia when it come to technology.