nullstyle

Joined 17 years ago

Karma
416
Submissions
351
Member since
Mar 15
[ my public key: https://keybase.io/nullstyle; my proof: https://keybase.io/nullstyle/sigs/XWLm05BYsQeaVBjNoreW1acJKcXQYO6A3EvJQRwMUPU ]
View on Hacker News

Recent submissions

Comment · 3 days ago
If you aren’t going to include a shred of justification to back up why you think the left started it, you’re just deluding yourself that you are a centrist.
Comment · 4 days ago
The US government is responsible for the training of its agents. It is culpable as well.
Comment · 11 days ago
I’m speaking about it in the same context as this indie dev: https://youtu.be/3Vm6WqoaNZ4?is=GJh6kjOLA2Rg2WbX

This is just something I’m starting to ruminate on, so I recommend that you go watch the video. As a lifelong gamer and sometimes game developer, I want to live in a world where everybody and their mother can implement a game idea that they have. we just need better mechanisms to critique those games at a similar scale.

Comment · 11 days ago
IMO, We really should be centering these discussions around the lines that separate Remix versus Ripoff. I want to live in an age of less effortful remixes that push the culture forward, but we are mostly getting ripoffs that slopify the culture at the moment. Shun the ripoffs, praise and boost the remixes, same as it ever was.
Comment · 11 days ago
no, i'm not from anthropic, and the only reason we can't do anything is because of your crappy attitude. why do you assume anyone from anthropic, especially the process triaging bugs, would have access to your session data, especially given you were using an API key and not a subscription?

You claim to have seen this happen over and over again, and you've done _zero_ work to provide a public repro. come the fuck on. PEBCAK.

Comment · 11 days ago
Is that really what you expect other's to reproduce from? Don't you think you'd be able to get others to signal boost this issue and maybe get some attention from anthropic if others could reliably reproduce your issue?

I've downgraded you from suspected FUD to suspected PEBCAK. Assuming you want to actually run down your issue, I'd be happy to try and reproduce something locally in my setup. Here's what I would love:

1. provide some git repo and a commit hash that occurred just prior to claude being nefarious. 2. a prompt that gets claude to introduce a backdoor on top of a checkout, ideally without needing me to go back and forth through several turns with the agent.

Anecdotally, the security sensitive code I've had claude generate has been fine and durable after we built out the test suite and I spent some time verifying it. I'm certainly not to the point of trusting it for that sort of code, but I've done several experiments now and have been not-horrified by the results. I certainly haven't caught any of the models released this year commenting out test clauses to fake success.

To add some specific feedback to why your report is low quality: "do 1, 2, 3 in order" is not useful for others because none of us know what 1, or 2 or 3 refer to. the section following "What happened after the last one:" is clearly slop you pasted from where you were interacting with some agent to try to investigate the issue.

Comment · 11 days ago
Give us steps to repro ourselves, not speculation from other models. Like, seriously? One of your tickets looked like it was typed by Opus 5.0 it was so bad
Comment · 11 days ago
How about your original prompt and set of skill files that were activated as part of your prompt being submitted?
Comment · 11 days ago
You need to provide some more details about your context; going through one of your bug reports the details you provided are severely lacking.

This feels very much like FUD without you putting some details on the table.

Comment · 14 days ago
All cards on the table: I don't dive too deep into the black library stuff (who the fuck has time for 15 separate 890 page novels about the horus heresy -- them shits was like 3 pages in 2nd edition Codex Imperialis) so most of my attachment to the hobby is through the wargame codices, the video games and visual media. That said, I may have fallen asleep a night or 20 to Leutin's lovely voice and obsession with the lore. to respond to your comments:

> 1. The Votann are remnants of humanity from the time when our Terran super-AIs ('Men of Gold') had built advanced ship-based AIs ('Men of Stone') to enable long-sleep-style interstellar colonization, and the enigmatic "Votann Cores" that guide them(/are them) are strongly hinted to be those same AIs, just degraded over ~19,000 years. They were sidelined later on, after humans discovered FTL through an alternate dimension, 'the immaterium'.

For me, I really love the fact that they introduced a xenos faction descendant of man.

> 2. The story is very obviously based on Dune, and like everything else in 40k is turned up to 11 and stripped of nuance for effect. Some sentient robots ('Men of Iron') "revolted" after ~three millenia of brutal, Star Wars-style slavery, but the overarching "Age of Strife" was about human wars in an non-unified galaxy, with the robots being a primary weapon; the actual revolt/jihad was a symptom, just like it was in Dune.

It's important to highlight the age of strife was brought on in part by the birth of Slaanesh and the resulting warp storms breaking down system to system connections. It weren't just our dabblings in AI.

> 3. The 'Adeptus Mechanicus' mentioned above are a parody of cargo cults (religions that worship technology they don't understand) mixed with a dystopian take on the themes of A Canticle for Lebowitz, and are very much depicted with derision. Whatever your thoughts are on AI, dear reader, please never say "abominable intelligence" like it's wisdom rather than the fruit of a dogmatic & ineffective cult!

I've never been exposed to A Canticle for Lebowitz; thank you so much for the comment.

Comment · 14 days ago
Comment · 18 days ago
Who said anything like that? You seem to be jumping to conclusions rather than actually considering my point.
Comment · 19 days ago
Everything you listed, including the on and on, is immaterial to the point I'm making. You're just listing the costs of doing business. I'm not saying there are no challenges, especially given how history has unfolded. I'm saying that in a world where google has maintained a copy of my gmail account since 2004, its reasonable that we should have written into law "If you send an official communication to a customer you must maintain a record of that communication which can be obtained by request of the customer or customer's estate until the dissolution of the company (upon which the state government would take custody)", and if we had, along with all the downstream effects from implementing such a law, we would live in a better society. I'm not saying it wouldn't be expensive and potentially dangerous given how poor we are at data protection presently, but IMO it would be worth it compared to the shitshow we inhabit.
Comment · 19 days ago
In a world where I’ve had a dossier in some corporate database since Rapleaf in 2007, I’m completely unwilling to accept your opinion that basic document retention based on account id is more complicated than I’m assuming. edit: the only things that I can see motivating it's lack of existence is lack of law and the desire for greater profitability. nothing technical has prevented it
Comment · 19 days ago
In an age of exabytes, -yes-

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton