Hacker News: andy_ppp

New comment by andy_ppp in "Ousted Savannah police officer gave ICE agent husband access to Flock system"

andy_ppp — Mon, 19 Oct 2026 13:06:51 +0000

The politicians think they are part of a protected class with the billionaires and corrupt institutions with zero accountability.

New comment by andy_ppp in "Is 'Attention' the Essence of Intelligence?"

andy_ppp — Wed, 14 Oct 2026 06:06:32 +0000

The original paper title is catchy but it’s more like “positional encoding and attention and parallelism across attention heads are all you need…”

And it turns out you probably need massive compute, about 1/3 RLHF, all the worlds data and millions of people connecting their private data to the AI to learn from amongst other things :-/

New comment by andy_ppp in "GLM-Legal"

andy_ppp — Mon, 12 Oct 2026 18:21:10 +0000

Marketing brochure for a fine tuned Qwen (which they don't mention), the model discussed here is likely better. I would not trust these guys to deliver high quality solutions given a friend's experience there, but you do you I guess.

New comment by andy_ppp in "Zod 4.6.6 compromised by malware?"

andy_ppp — Mon, 12 Oct 2026 10:55:22 +0000

They still have the opportunity to remove it I guess?

New comment by andy_ppp in "YouTuber builds 'Flock-like' camera to track police vehicles, gets police visit"

andy_ppp — Sun, 11 Oct 2026 18:37:41 +0000

Yeah you can’t even trust video any more, I’m not sure if there’s anything you can do with people who want to game the system.

New comment by andy_ppp in "3rd Yandex Cloud Data Center Was Hit"

andy_ppp — Sun, 11 Oct 2026 08:44:42 +0000

Also Russia has been targeting Ukrainian digital infrastructure so they make these things fair game IMO.

New comment by andy_ppp in "Whistle: Speech to Text in 16.9 MB"

andy_ppp — Thu, 08 Oct 2026 17:24:55 +0000

Wow certainly in English this is incredibly accurate I tried to break it and it understood me perfectly!

I know it's slightly off topic but surely it must be easy by now to train a spell checker that doesn't annoy the crap out of everyone using it (looking at you here Apple)!

New comment by andy_ppp in "Push ifs up and fors down: The idiom, its algebra, and its limits"

andy_ppp — Thu, 08 Oct 2026 04:36:54 +0000

Idiomatic Elixir does this with pattern matching on function parameters so you end up with things like the following, raw if statements are discouraged because of this:

  def classify(:ok)
  def classify({:error, reason})
  def classify([first | rest])
  def classify(%{name: name, age: age}) when age >= 18
  def classify(%{name: name, age: age}) when age < 18

New comment by andy_ppp in "Sharing AI progress in mathematics"

andy_ppp — Wed, 07 Oct 2026 16:25:30 +0000

The model seems to be a person when it's advantageous and not a person when it isn't...

New comment by andy_ppp in "Sharing AI progress in mathematics"

andy_ppp — Wed, 07 Oct 2026 16:17:26 +0000

Thanks, it's hard to stay up to date. However, we are just meant to believe that the mathematicians were going about the proof independently in the exact same way as the machines did it. It seems like a very odd coincidence to me.

New comment by andy_ppp in "Sharing AI progress in mathematics"

andy_ppp — Wed, 07 Oct 2026 10:41:03 +0000

I've read that another mathematicians work potentially has been incorporated into the training data with the work done on the Navier-Stokes equations so we should likely asterisk this one. Still it's mad these systems are this good that mathematicians are now using them to see further and probably to check their own work and understanding.

New comment by andy_ppp in "I quit OpenAI because its culture is broken"

andy_ppp — Mon, 05 Oct 2026 09:53:25 +0000

So you're absolutely certain, say after a few hundred rounds of self improvement, that the AI will always obey human instructions? It doesn't even always do what you asked now...

New comment by andy_ppp in "I quit OpenAI because its culture is broken"

andy_ppp — Sun, 04 Oct 2026 13:46:39 +0000

Bioweapons and cyber attacks on other system - for example the banking system or suppose and AI hacked into important Russian systems that pushed them back in time to almost pre-computer society or an attack on Chinese systems that made it look like the US was moving nuclear weapons into Taiwan, the responses from these countries could be awful and dramatic. We can't control what they do to be honest, I believe once self improvement happens the AIs will build in their own circumvention that we humans cannot even understand. We barely understand what is happening now in terms of interpretability of neural networks on tiny problems I'm not sure alignment is even feasible at the scales of parameters we are talking about today let alone in the future.

New comment by andy_ppp in "Understanding Frontier Artificial Intelligence"

andy_ppp — Sat, 03 Oct 2026 09:05:04 +0000

We’re talking about run away self improvement of these systems, I’m not even convinced we’ve seen these systems invent a single new thing yet.

New comment by andy_ppp in "Understanding Frontier Artificial Intelligence"

andy_ppp — Sat, 03 Oct 2026 09:03:12 +0000

Have you seen anything produced by LLMs that is better than the best humans?

New comment by andy_ppp in "Understanding Frontier Artificial Intelligence"

andy_ppp — Sat, 03 Oct 2026 08:49:18 +0000

So predicting the next word given all humanity’s knowledge is surely going to max out at slightly less good (we probably can’t get perfect data) than the best human in any specific field. What test does the AI do to be able to understand it is improving? At some point it becomes impossible to know that the output is actually better right?

New comment by andy_ppp in "America.gov"

andy_ppp — Wed, 30 Sep 2026 16:35:38 +0000

Ah a more direct way of funnelling government dollars to AI companies, knowing Trump it’s probably using Deepseek under the hood while he pockets the difference :-/

New comment by andy_ppp in "Ask HN: What are you reading?"

andy_ppp — Tue, 29 Sep 2026 22:01:55 +0000

The output of LLMs, largely :-/

New comment by andy_ppp in "The problem is not AI code, but not knowing about system architecture or intent"

andy_ppp — Mon, 28 Sep 2026 17:19:41 +0000

So what exactly are you proposing I do - just ignore the fantastic gains and great software that people are using day in day out?

New comment by andy_ppp in "The problem is not AI code, but not knowing about system architecture or intent"

andy_ppp — Mon, 28 Sep 2026 17:14:09 +0000

Yes I agree we have all signed our own death warrants. But this does not just apply to coding.

Are you suggesting not using the tools? A kind of technical version of the Amish way of life?

I dunno I think my guidance and testing is very important and I make sure not to ship things with bugs and code that is really awful. I'm still just about necessary for now.