HNHacker News
TopNewBestAskShowJobs

chrisjj

3,805 karma · joined July 11, 2014

submissionscomments
chrisjj··on Superlite: A Faster Alternatie to Superpowers
> You unambiguously told it what you wanted, it did that

Try rereading the report please.

chrisjj··on Superlite: A Faster Alternatie to Superpowers
These bot are real smart.
chrisjj··on Superlite: A Faster Alternatie to Superpowers
> Since then I’ve been back to vibing. I lose overnight runs but gain sanity.

You gain sanity only if (insanely) you never test the output.

chrisjj··on Superlite: A Faster Alternatie to Superpowers
> The models are probabilistic in their output.

The word you're looking for is unreliable.

> But because the models are probabilistic, the real problem is that it is guaranteed that some of the places would not be updated

untrustworthy.

chrisjj··on Superlite: A Faster Alternatie to Superpowers
> The code is 'correct', but it's just not what any of us would have written".

I'd fine with not what we'd have written ... if it worked.

Trying ChatGPT for a simple one page of HTML JS having draggables, I said clamp touch point to visible viewpoint. Fine. Then I said add 3x drag gain. It put that after the clamp, whereas anyone/thing having an ounce of intelligence would have urecognised the only possible purpose of the clamp and so put the gain before.

This is fundamentally why we cannot trust stochastic parrots to code. They lack the human's understanding of purpose which cannot be abstracted from the code yet is essential for success.

chrisjj··on Superlite: A Faster Alternatie to Superpowers
> However is fastest, please, I beg you.

Fast is easy.

Fast and good, not so much.

chrisjj··on NY announces game ownership bill requiring ownership rather than licensing
"GOG content" is the software.
chrisjj··on NY announces game ownership bill requiring ownership rather than licensing
That transfer would breach the term prohibiting transfer.
chrisjj··on Google Meet Enshittification
> You mean you need permission to show notifications to join a meeting?

I mean it claims so. You click a join link, it says notification permission needed.

> I think it is needed otherwise the system would kill the app and the call as soon as you either leave the app (which is possible while having an ongoing call)

Not reproduced here. Samsung phone.

chrisjj··on NY announces game ownership bill requiring ownership rather than licensing
You'd be in breach of the terms.
chrisjj··on NY announces game ownership bill requiring ownership rather than licensing
Technically legally no.

"Your GOG account and GOG content are personal to you and cannot be shared with, sold, gifted or transferred to anyone else."

chrisjj··on What's the management thinking behind enshittification
Such outcomes used to disavantage the supplier, leading to elimination by natural selection.

Less so now. It seems the selection rules have changed somehow.

chrisjj··on What's the management thinking behind enshittification
And what caused this change do you think?
chrisjj··on What's the management thinking behind enshittification
But then why has this become prevalent only recently?
chrisjj··on What's the management thinking behind enshittification
I think it wasn't previously.

And enshittification is rarely less plastic or anything, but more ads etc.

chrisjj··on How the Claude Code team does code review
> 'Now is the cheapest time it will ever be to experiment'?

First hit is free ;)

chrisjj··on How the Claude Code team does code review
> The agent will fix a bug when I tell it to, but only in exactly one place. I'll notice that another place has that bug later and have to redo it. It's the sort of thing that I would know to check for if I were fixing it myself.

I see the same in e.g. simple CSS. I ask ChatGPT to apply styling X to aĺl instances of say a UI control type. Then to another. Later I ask for a tweak of X. Bot updates some but not all instances. It didn't create a single class for all. No intelligence seen.

The most remarkable fails are simpler though. E.g. at version 303 I ask "Revert to version 299". I get 299 with 303's edits. Bot explains it interpreted "Revert" to mean something no intelligent programmer would.

And most importantly, these are just the ones I directly notice. To see the total, I have to read through and understand the most repulsive code I've seen in my life.

The only way to be satisfied with bot code is by having exceptionally low expectations. And never ever let yourself be kidded that your helper is more than a stochastic parrot gorged on stackoverflow etc.

chrisjj··on How the Claude Code team does code review
> One thing I do is ask the LLM to write me a walkthrough of the code.

I tried that. ChatGPT's result was laden with loud, clear and wrong fabications which I might have believed had I not examined the code.

Unreliable program be unreliable.

chrisjj··on How the Claude Code team does code review
> Most people who are impressed by it, I am sure, are just shit coders themselves

The less you understand code, the more you're impressed by coding chatbots - in part because you're less likely to read the bot's code, and if you do, less likely to realise how poor it is. Code gets judged good simply because it exists and costed little to create.

chrisjj··on What's the management thinking behind enshittification
Accepted, but why only recently has that loaded signal pointed shitwards?
chrisjj··on What's the management thinking behind enshittification
But that did not previously entail shit level must go up.
chrisjj··on Japanese people still aren't having sex. And nobody knows why
> And this national dry spell seems to have got worse

Hmm. Did the subjects say worse?

chrisjj··on Why We're Banning AI
> I use LLMs for fast feedback to see if my words are having the intended effect (i.e., are they inducing emotions?

Why? Your chatbot has ha no more experience of emotions than your teapot.

chrisjj··on Why We're Banning AI
> Nothing in our rules ... artist from using a tool

So "AI" tool OK?

> to, say, add a texture or change a color, or make some kind of small element.

So "AI" OK for elements of some value of "smaller"?

This ban sounds somewhat ... flexible.

chrisjj··on Neural networks are clean
> After many months studying neurons and circuits in image and language models

There are no neurons or anything like them in these image and language models.

chrisjj··on A.I. Companies Say They Aren't Responsible for Their Unpredictable Products
Gift link, anyone?
chrisjj··on A.I. Companies Say They Aren't Responsible for Their Unpredictable Products
Chatbot providers: "Our AIs are more intelligent than us!"

Yes. Chatbots excel at leaping low bars.

chrisjj··on A.I. Companies Say They Aren't Responsible for Their Unpredictable Products
True title: A.I. Companies Say They Aren’t Responsible for Their Unpredictable Products. Don’t Believe Them.
chrisjj··on Ledger confirms spy chip implant behind wallet crypto thefts
> there is no authority that can undo or correct illegally made transactions

That is the only way to ensure no "authority" can undo or correct legally made transactions.

chrisjj··on Redesigning source control at Anthropic
How can I put this politely ...

> I rarely write code anymore. The code that I get from the LLM is just better.

Coding chatbots excel at leaping low bars.

← PreviousPage 4 of 34Next →