Hacker News: Turskarama

New comment by Turskarama in "The other half of AI safety"

Turskarama — Thu, 14 May 2026 01:18:03 +0000

Just because the users were already sick when they started using ChatGPT doesn't mean that ChatGPT isn't exacerbating the issue. Sickness isn't a boolean condition. A big problem with LLMs in general when it comes to people like this is that they are too sycophantic, they don't push back when you start acting strange and they're too gentle about trying to validate you.

New comment by Turskarama in "Vibe coding and agentic engineering are getting closer than I'd like"

Turskarama — Thu, 07 May 2026 04:12:27 +0000

In my experience AI can write _something_ from scratch, but often edge cases won't be handled until I go through and read the results or test it. Usually when I'm writing by hand I will naturally find the majority of edge cases as I go. By the time I've read through the results and fixed said edge cases, I usually would have been faster just doing it myself.

New comment by Turskarama in "What I'm Hearing About Cognitive Debt (So Far)"

Turskarama — Tue, 05 May 2026 05:31:14 +0000

Agile cannot go over budget or scope because those are failures of planning. Agile is the methodology that was developed specifically to counteract those problems with planning. Projects that use Agile can go over budget and scope but they never do that because they are using Agile, rather they use Agile because they might do that.

New comment by Turskarama in "The Joy of Folding Bikes"

Turskarama — Mon, 27 Apr 2026 08:35:20 +0000

Which is fine, the average bike thief is not going to expect to need them.

New comment by Turskarama in "The peril of laziness lost"

Turskarama — Mon, 13 Apr 2026 00:16:26 +0000

Yes, there's a reason it was abandoned as a KPI almost as soon as it was introduced. Just because AI is writing the code instead now doesn't magically make it a good metric.

New comment by Turskarama in "Union types in C# 15"

Turskarama — Thu, 09 Apr 2026 08:26:14 +0000

You should have kept reading.

  For performance-sensitive scenarios where case types include value types, libraries can also implement the non-boxing access pattern by adding a HasValue property and TryGetValue methods. This lets the compiler implement pattern matching without boxing.

New comment by Turskarama in "Our commitment to Windows quality"

Turskarama — Sat, 21 Mar 2026 07:50:24 +0000

If Windows update does something weird what do you do?

New comment by Turskarama in "The American Healthcare Conundrum"

Turskarama — Tue, 17 Mar 2026 11:49:35 +0000

> In many ways the quality of care in the US is far better than what folks get elsewhere, which in part is probably why there isn’t a total patient rebellion

How sure of this are we really? Other countries mostly have problems with emergency departments being full, but that's less because those emergency departments are worse and more because in the US people aren't going, they just stay home and hope they don't die.

New comment by Turskarama in "My “grand vision” for Rust"

Turskarama — Mon, 09 Mar 2026 13:41:50 +0000

Actually you're right, I was going by the source code which was in the link of the comment you replied to, but I missed that that was specifically for divExact and not just primitive division.

New comment by Turskarama in "My “grand vision” for Rust"

Turskarama — Mon, 09 Mar 2026 11:50:09 +0000

More specifically, Zig will return an error type from the division and if this isn't handled THEN it will panic, kind of like an exception except it can be handled with proper pattern matching.

New comment by Turskarama in "AI isn't "just predicting the next word" anymore"

Turskarama — Tue, 13 Jan 2026 23:51:57 +0000

Have you ever had a concept you wanted to express, known that there was a word for it, but struggled to remember what the word was? For human thought and speech to work that way it must be fundamentally different to what an LLM does. The concept, the "thought", is separated from the word.

New comment by Turskarama in "The insecure evangelism of LLM maximalists"

Turskarama — Tue, 13 Jan 2026 23:45:43 +0000

I have seen people say something along these lines what feels like every month for the past year.

New comment by Turskarama in "Gemini 3 Flash: Frontier intelligence built for speed"

Turskarama — Thu, 18 Dec 2025 00:28:43 +0000

The point is that it's a litmus test for how well the models do with niche knowledge _in general_. The point isn't really to know how well the model works for that specific niche. Ideally of course you would use a few of them and aggregate the results.

New comment by Turskarama in "Working quickly is more important than it seems (2015)"

Turskarama — Thu, 18 Dec 2025 00:08:19 +0000

I think ideally you need to practice both slow AND fast. You need to practice slow so you can notice and work on small details that can be skipped over with speed, and you need to practice fast because some things are legitimately different at speed and you won't learn how to deal with them only going slow.

New comment by Turskarama in "Gemini 3 Flash: Frontier intelligence built for speed"

Turskarama — Wed, 17 Dec 2025 22:46:00 +0000

The problem with publicly disclosing these is that if lots of people adopt them they will become targeted to be in the model and will no longer be a good benchmark.

New comment by Turskarama in "No AI* Here – A Response to Mozilla's Next Chapter"

Turskarama — Wed, 17 Dec 2025 02:22:13 +0000

The problem with integrating a chat bot is that what you are effectively doing is the same thing as adding a single bookmark, except now it's taking up extra space. There IS no advantage here, it's unnecessary bloat.

New comment by Turskarama in "Economics of Orbital vs. Terrestrial Data Centers"

Turskarama — Tue, 16 Dec 2025 07:20:31 +0000

The computer chips used for AI generate significantly more heat than the chips on the JWST. The JWST in total weighs 6.5 tons and uses a mere 2kw of power, which is the same as 3 H100 GPUs under load, each of which will weight what, 1kg?

So in terms of power density you're looking at about 3 orders of magnitude difference. Heating and cooling is going to be a significant part of the total weight.

New comment by Turskarama in "Helldivers 2 devs slash install size from 154GB to 23GB"

Turskarama — Thu, 04 Dec 2025 04:43:41 +0000

They didn't actually fix this until a couple of months after they publicly revealed that this was the reason the game was so big and a lot of people pointed out how dumb it is. I saw quite a few comments saying that people put it on their storage HDD specifically because it was too big to fit on their SSD. Ironic. They could have got their own data quite a bit earlier during development, not nearly two years after release!

New comment by Turskarama in "Everyone in Seattle hates AI"

Turskarama — Thu, 04 Dec 2025 04:24:03 +0000

It actually is somewhat a limit of the technology. LLMs can't go back and modify their own output, later tokens are always dependent on earlier tokens and they can't do anything out of order. "Thinking" helps somewhat by allowing some iteration before they give the user actual output, but that requires them to write it the long way and THEN refactor it without being asked, which is both very expensive and something they have to recognize the user wants.

New comment by Turskarama in "I Stopped Being a Climate Catastrophist"

Turskarama — Thu, 06 Nov 2025 12:57:28 +0000

There is one absolutely massive one, and that's that for the first time the problem is truly global. Other famines have been caused either by war or local droughts, both of which affect only a population in a limited area and crucially, can be somewhat mitigated by importing food from elsewhere. You can't import food if there are global food shortages.