Jan
Immortal WombatINTERCAL: We have keywords like PLEASE and IGNORE because we think it's funny
Anthropic: We have PLEASE and IGNORE *baked into our source-code* because we have no fucking idea how this thing works
Anthropic: "You have no idea what's coming. And neither do we. We literally have no idea how any of this works. Our source-code looks like something you'd be embarrassed to present as a GCSE assignment"
This shambolic dogshit is coming for your white-collar job in 12-18 months*
At least the machine looms fucking WORKED
*12-18 month rolling window, as ever
When this all comes crashing down, remember how little the Epstein Class valued you and your labour
Remember also how little some people seem to care about quality and correctness
Half-assing it at scale
Patrick@pikesley I remember when they decided that it was better to boil the Earth and everyone living on it instead of lower consumption just even a little bit. Seems rather par for the course right now.
grob (teeth era) πΊπ¦π³οΈβππ³οΈββ§οΈ@PatrickoftheG @pikesley I really need to use "par for the course" more often. It is such a good phrase with indifferent disdain baked into it 
The Orange Theme@pikesley And 90% of my colleagues, apparently.
Mrs.Malice β€οΈβπ₯β¨@theorangetheme @pikesley Oof, the colleagues. That part right there has hurt the most in all this, at least for me.
esaul
Emma: Ζ(βοΈ, πββ¬οΈ, ποΈ) β β@pikesley AGI and fusion are always 20 years away.
@emma and a cure for my Type-1 Diabetes (although that's sometimes 10 years)
Cap Ybarra@pikesley goal posts have shifted, it now doesn't work but that's ok and things have always not worked and you should keep paying the same amount of money anyways
Andrea (Drea) Tamar Pinski@pikesley There was no GCSE for computer science when I was in secondary school :(. If there was I definitely would have taken it.
4 years after my GCSEs (well I only got a maths one but that is a different story) I was contributing to GCC even.
4 years after my GCSEs (well I only got a maths one but that is a different story) I was contributing to GCC even.
@robinadams @pinskia I have a GCSE in Computing from 1990 (we were the third GCSE cohort I think)
@pikesley @robinadams So maybe the secondary school I went to didn't offer it at that time which was expected from a school is a poor working class neighborhood of Coventry. They had acorn machines though which gave me a good introduction into that :).
Jez Higgins@pinskia @pikesley @robinadams There were CSEs in Computer studies, which involved a decent amount of project work) and O levels in Computer Science (which I think were just exams) prior to to the GCSE.
No idea how widely available they were.
The CSE was pretty easy. I got a grade 1 with things I learned from Your Computer and Popular Computing Weekly magazines. (I'm not exaggerating - I did it as an extra off timetable, so did no lessons)
* Serve the public trust
* Protect the innocent
* Uphold the law
* \u2014
* Protect the innocent
* Uphold the law
* \u2014
I'm now hearing where it's important to be polite to the idiot robot because then you might get better results CAN YOU FUCKING HEAR YOURSELVES?
samiamsam
Sean Middleditch@pikesley Since we have no idea what the training data actually is, there's a chance it *does* give better results, for some definition of "better" anyway.
Someone should do a study on the quality of results between asking tersely worded questions, asking politely worded questions, and burning the entire industry to the ground.
For science.
Henryk PlΓΆtz@pikesley Wait, last year's best magical practices was to *not* say please and thank you because it was costing them millions and making them sad.
@henryk well the cutting-edge Applied Magical Thinking now suggests we have to be nice
I hate it here
EndlessMason@pikesley
"Might"
"Might"
Kevin Granade@pikesley this is very much a problem with gambling based outputs. Every time you change your prompt it's a new pull on the slot machine for a "good" output. As such it gives biases maximum opportunity to kick in, since you also get a new pull on repeat prompts.
It's actually in the vendors interest to keep changing recommendations so that people stay in the, "you're promoting it wrong, just keep making 'corrections' until you like the output" loop. Shifts the blame to the user.
Artemis@kevingranade @pikesley I think this is also where all those "You're a senior <insert technology here> developer, please obey all the best practices" stuff comes from.
Amo Bishop Rodent (@[email protected])
The notion of "best practice" is predicated upon the ludicrous idea that anybody knows what they're doing
John Colagioia@pikesley It occurred to me that "not knowing how any of this works" is ALSO baked into the process, because you can't sell the fantasy of AGI on semantic vectors and linear algebra.
Apprentice Bodega Cat πββ¬@pikesley point me to more people dunking on Claude's source code, plz
I need the schadenfreude
@LeDiva @pikesley This guy dives into the details. https://neuromatch.social/@jonny/116324676116121930
jonny (nonvenomous) (@[email protected])
- Claude code source "leaks" in a mapfile - people immediately use the code laundering machines to code launder the code laundering frontend - now many dubious open source-ish knockoffs in python and rust being derived directly from the source What's anthropic going to do, sue them? Insist in court that LLM recreating copyrighted code is a violation of copyright???