Architeuthis

joined 2 years ago
[–] Architeuthis@awful.systems 7 points 4 hours ago

It's like a one-and-a-half-page article that also comes in audio and video form, don't be lazy.

spoilerThey vibe coded a bash injection vulnerability in their devops code, which was used to gain access to the repo and push out a release with malicious code, which prompted any installed LLM wrappers like cursor to gather anything that looked like a configuration or text file in the infected machine and presumably leak them to the attacker.

Modern move money between pockets for profit economics seem to give The Hitchhiker's Guide bistromathics a run for their money.

[–] Architeuthis@awful.systems 5 points 2 days ago* (last edited 2 days ago) (1 children)

I wonder what this means for US GDP

Don't worry, unchecked inflation and increasing housing costs will keep the GDP propped up at least for a while longer.

[–] Architeuthis@awful.systems 7 points 2 days ago (1 children)

Zitron taking every opportunity to shit on Scott's AI2027 is kind of cathartic, ngl

[–] Architeuthis@awful.systems 7 points 2 days ago (1 children)

He has capital L Lawfulness concerns. About the parent and the child being asymmetrically skilled in context engineering. Which apparently is the main reason kids shouldn't trust LLM output.

Him showing his ass with the memory comment is just a bonus.

[–] Architeuthis@awful.systems 11 points 3 days ago (5 children)

I feel dumber for having read that, and not in the intellectually humbled way.

[–] Architeuthis@awful.systems 14 points 3 days ago* (last edited 3 days ago) (22 children)

This hits differently over the recent news that ChatGPT encouraged and aided a teen suicide.

transcriptKelsey Piper xhitted: Never thought I'd become a 'take you relationship problems to ChatGPT' person but when the 8yo and I have an argument it actually works really well to mutually agree on an account of events for Claude and the ask for its opinion

I think she considers the AIs far more knowledgeable than me about reasonable human behavior so if I say something that's no reason to think it's true but if Claude says it then it at least merits serious consideration

[–] Architeuthis@awful.systems 4 points 4 days ago

AI innovation in this space usually means automatically adding stuff to the model's context.

It probably started meaning the (failed) build output got added in every iteration, but it's entirely possible to feed the LLM debugger data from a runtime crash and hope something usable happens.

[–] Architeuthis@awful.systems 13 points 6 days ago (1 children)

When I was at computer toucher school at about the start of the century, under the moniker AI were taught (I think) fuzzy logic, incremental optimization and graph algorithms, and neural networks.

AI is a sci-fi trope far more than it ever was a well-defined research topic.

[–] Architeuthis@awful.systems 1 points 6 days ago

Anyone who said this about their product would almost certainly by lying, but these guys are extra lying.

For sure, blockchain based agentic LLM that learns as it goes is sounds like someone describing a flying elephant wearing an inflatable life jacket.

 

Sam Altman, the recently fired (and rehired) chief executive of Open AI, was asked earlier this year by his fellow tech billionaire Patrick Collison what he thought of the risks of synthetic biology. ‘I would like to not have another synthetic pathogen cause a global pandemic. I think we can all agree that wasn’t a great experience,’ he replied. ‘Wasn’t that bad compared to what it could have been, but I’m surprised there has not been more global coordination and I think we should have more of that.’

view more: next ›