this post was submitted on 23 Jun 2024
34 points (100.0% liked)

TechTakes

1430 readers
118 users here now

Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.

This is not debate club. Unless it’s amusing debate.

For actually-good tech, you want our NotAwfulTech community

founded 1 year ago
MODERATORS
 

Need to make a primal scream without gathering footnotes first? Have a sneer percolating in your system but not enough time/energy to make a whole post about it? Go forth and be mid: Welcome to the Stubsack, your first port of call for learning fresh facts of Awful you'll near-instantly regret.

Any awful.systems sub may be subsneered in this subthread, techtakes or no.

If your sneer seems higher quality than you thought, feel free to cut’n’paste it into its own post — there’s no quota for posting and the bar really isn’t that high.

The post Xitter web has spawned soo many “esoteric” right wing freaks, but there’s no appropriate sneer-space for them. I’m talking redscare-ish, reality challenged “culture critics” who write about everything but understand nothing. I’m talking about reply-guys who make the same 6 tweets about the same 3 subjects. They’re inescapable at this point, yet I don’t see them mocked (as much as they should be)
Like, there was one dude a while back who insisted that women couldn’t be surgeons because they didn’t believe in the moon or in stars? I think each and every one of these guys is uniquely fucked up and if I can’t escape them, I would love to sneer at them.

(page 2) 50 comments
sorted by: hot top controversial new old
[–] [email protected] 11 points 5 months ago (1 children)

seemingly yet more chatgpt jailbreaks just by providing input that barrierbreaks some n times, and then readily provides details

y'know, if I were the one making safety nets for these systems, I'd make them return such kind of results (or other typical honeypot type behaviour). and it's possible that that's what oai did. but it seems extremely unlikely that that's what they did, because it goes again the bayfucker gottagofast philosophy (and, frankly, against the level of competence I've seen displayed in the genml space overall)

[–] [email protected] 11 points 5 months ago* (last edited 5 months ago)

It turns out the 'I' in "AI" stood for "Linux" all along!

User friendliness aside (who in their right mind would want arbitrary code execution except shitty and indeterministic?), I sandbox stuff at my job* and it's hard to evaluate how secure / privacy preserving this is without more details.

If they're running a full fledged VM and super extra careful around the sandbox boundary** it's probably fine; otherwise it seems perhaps a bit loosey-goosey.

Someone will eventually try to run a Monero cryptocurrency miner in it if they haven't already. So I hope they have their timeouts and resource limits in order (actually I hope they don't, for the lols).

* But like no one told me how to do it or gave me a certificate or anything I just had to do my best

** This is often way scarier than programmers are used to, unless they've written a secure parser before. I wrote a vulnerability into my code a few years back when I was younger and foolish, by trusting an array length from inside the sandbox. My coworker found it while fuzzing the code.

[–] [email protected] 11 points 5 months ago (1 children)

surprise lemmy feature discovered:

[–] [email protected] 10 points 5 months ago (1 children)

oh good lord. we went live a year ago

[–] [email protected] 10 points 5 months ago

oh hell. we’re beating all my initial survivability projections by a lot

do we throw an instance birthday party thread? will there be cocktails? will the deployment get mopey if I don’t buy it more disk space? (yes, eventually)

load more comments
view more: ‹ prev next ›