Hacker News: briansm

New comment by briansm in "AMD acquires Taalas to boost inference performance by etching models in silicon"

briansm — Sat, 08 Aug 2026 05:31:07 +0000

Cerebras have special techniques to work around etching errors / bad cores on their wafers. This is possible since their wafers are effectively hundreds of identical copies of redundant cores. Can't do that for a globally unique model.

Etching failure in that situation would be like brain-damage in a human, all sorts of weird effects would start appearing.

New comment by briansm in "There Will Come Soft Rains (1950) [pdf]"

briansm — Tue, 04 Aug 2026 07:13:42 +0000

Reminds me of when November 2019 came and went (the date that Blade Runner is set in)

New comment by briansm in "ReFrame – The EPaper Camera"

briansm — Wed, 29 Jul 2026 09:29:26 +0000

Make hay while the sun shines.

New comment by briansm in "Data Compression Explained (2012)"

briansm — Sat, 20 Jun 2026 07:49:58 +0000

I think so, specifically lossy compression though.

A modern version of the book would include an extra section in the 'Lossy compression' chapter - 'Text' (alongside Images/Video/Audio) that would discuss LLM's.

New comment by briansm in "πFS"

briansm — Thu, 11 Jun 2026 11:36:41 +0000

LLM's seem to be the weird interesting outcome of applying lossy (de)compression concepts to text instead of the audio/image/video domains where they have traditionally been used.

New comment by briansm in "Gemma 4 12B: A unified, encoder-free multimodal model"

briansm — Thu, 04 Jun 2026 11:40:14 +0000

Strange that they are feeding raw audio in. Even in humans, there is a hardware transform to the frequency domain (the cochlea) before data is fed to the brain, effectively doing this part in the LLM seems inefficient.

New comment by briansm in "WriteUp: 16 Bytes of x86 that turn Matrix rain into sound"

briansm — Mon, 18 May 2026 09:14:54 +0000

My favorite is 'spongy', an underwater journey through a Menger-sponge fractal in 128 bytes:

https://www.pouet.net/prod.php?which=53871

https://www.youtube.com/watch?v=36BPql6Nl_U

I'm more impressed by the eerie beauty of it than the technical achievement, even if the code was orders of magnitude larger it would still be wonderful.

New comment by briansm in "WriteUp: 16 Bytes of x86 that turn Matrix rain into sound"

briansm — Mon, 18 May 2026 07:03:28 +0000

For a more laid-back chill 'rain' demo, have a look at "Rainy 32b" 32-byte x86 demo:

https://www.pouet.net/prod.php?which=86923

https://www.youtube.com/watch?v=iCgIQuPHb5w

New comment by briansm in "I turned a $80 RK3562 Android tablet into a Debian Linux workstation"

briansm — Mon, 18 May 2026 05:22:49 +0000

I had to read that twice, I first thought you meant 4 GB of _storage_...

Worrying about only having 4GB of RAM.... that's how I know I'm getting old I guess. Kids these days don't know they're born.

New comment by briansm in "Colossus: The Forbin Project"

briansm — Sun, 17 May 2026 11:31:12 +0000

I've always wondered if it was perhaps the inspiration for the novel Neuromancer (2 AI's in different continents plotting to combine with each other to form a global super-intelligence)

New comment by briansm in "US inflation jumps to 3.8% as energy costs surge from Iran war"

briansm — Tue, 12 May 2026 16:00:06 +0000

We got both, so twice as good.

New comment by briansm in "US inflation jumps to 3.8% as energy costs surge from Iran war"

briansm — Tue, 12 May 2026 15:40:12 +0000

Ah, 'demand destruction', which in sounds positive in theory but in practice will likely mean poverty, famine and 'population reduction'.

New comment by briansm in "Removable batteries in smartphones will be mandatory in the EU starting in 2027"

briansm — Mon, 04 May 2026 22:05:53 +0000

You are assuming that a battery is full at 4.2V and empty at 0V. In practice a Lithium Ion battery is empty at 3V. You are also assuming the relationship between capacity and voltage is linear - it is not.

New comment by briansm in "New statue in London, attributed to Banksy, of a suited man, blinded by a flag"

briansm — Mon, 04 May 2026 21:51:21 +0000

In the developed world we have 'energy slaves' instead; around 100 per person in the USA in the 2020's.

https://en.wikipedia.org/wiki/Energy_slave

This is arguably the reason why the Overton window has shifted towards the rejection of human slavery over the last century or so, with the growth of fossil fuel use.

Human slavery will thus likely swing back into fashion again in the future as oil, coal and natural gas run out.

New comment by briansm in "US and Iran agree to provisional ceasefire"

briansm — Wed, 08 Apr 2026 09:03:07 +0000

That will happen on the day there is enough to go around.

Unfortunately I fear that time has long gone.

As long as the population is growing and/or the (fossil) energy is falling there will never be enough to go around.

New comment by briansm in "Department of State advises Americans worldwide to exercise increased caution"

briansm — Mon, 23 Mar 2026 07:53:38 +0000

I keep reading this, is it meant to be a joke? What is being won?

New comment by briansm in "Global warming has accelerated significantly"

briansm — Fri, 06 Mar 2026 15:36:16 +0000

People are losing their minds at the prospect of oil availability dropping just 20% for a month or two with the closing of the Strait of Hormuz - even just this could collapse the global economy.

So yea, no way is oil stopping or even dipping slightly any time soon.

New comment by briansm in "The Xkcd thing, now interactive"

briansm — Tue, 03 Mar 2026 13:14:19 +0000

Just to mention the original was cited in the most recent Veritasium video:

"The Internet Was Weeks Away From Disaster and No One Knew"

https://www.youtube.com/watch?v=aoag03mSuXQ

(at about the 9:50 mark)

New comment by briansm in "RAM now represents 35 percent of bill of materials for HP PCs"

briansm — Thu, 26 Feb 2026 07:56:29 +0000

Economies die from lack of produce though.

New comment by briansm in "How Taalas “prints” LLM onto a chip?"

briansm — Sun, 22 Feb 2026 10:20:29 +0000

I wonder if you could use the same technique (RAM models as ROM) for something like Whisper Speech-to-text, where the models are much smaller (around a Gigabyte) for a super-efficient single-chip speech recognition solution with tons of context knowledge.