some of the greatest programming books in the world have been written with Pascal as their language.

I sometimes use Gemini and then correct with Sonnet.

however, it still feels like a lot of work. what if AI hallucinates during the check?

Comments URL: https://news.ycombinator.com/item?id=50138528

Points: 2

# Comments: 3

It just seems so much work. Pascal was a breath of fresh air.

And by far the best GOL I've seen so far.

- re: testability, clearly you haven't worked deeply with AI. AI makes unpredictable mistakes. Can you test those? You can't.

- If y then x is not a real argument. You're saying [if x] but x can never be true. So that's a fallacious standpoint.

- LLMs don't 'reason'. They're just using sophisticated pattern-matching. Human error rate will never be beaten, as far as the tasks involve unpredictability (so something like chess is predictable). -

And I say this as someone who can comfortably read Kant in (the original no less).

Types of errors, and then the frequency as well. And then there are so many tools that help humans bring down quantity and seriousness of the errors. These tools were designed for those kinds of human errors (which are somewhat predictable). But you can't use them on the unpredictable AI errors.

At other times, AI is not strictly in error, but the outcome is not what you exactly wanted. Can you imagine this in a medical scenario?

Do you want to fly in a plane that goes 100% on Auto-pilot + AI (the best you have in say, 2045) versus a plane piloted by a competent and experienced human pilot? Imagine you're traveling with all of your kids and extended family members. Would you take that risk knowing that humans also make mistakes as you said? Which risk would you be comfortable taking?

Ai's improvement will never reach 100%-error-freeness. Those who work in the trenches know that (hint: heuristics-based). He claims as if AI improvement is algorithmics-based. Which it will never be.

At best, AI will assist in cases where human involvement might be tedious or (or risky, such as Japan sending those bots in radioactive waters post the quake). And even that, under human supervision.

which also makes me wonder if it wouldn't have been more programmer-friendlier had it been written in Perl?

By the by, for a Russian writer of Trollope caliber, I recommend M. Lermontov.

Wow, I didn't know Wertherian was an adjective! To be sure, Trollope himself or someone he knew may have experienced these feelings which show themselves in many different archetypes and due to different causes: in Mr. Gibson, in John Eames, in Adolphus Crosbie, etc. Sometimes it also appears in women characters.

I am now on a Trollope binge. That man distilled all of his hard life experiences as musings interspersed between the social dilemmas his characters were facing.

Recommend you start with either the "Eustace Diamonds" or "The small house at Allington".