RatFace5000
@lurg.bsky.social
Timeline where companies comply and their LLMs start putting vulnerabilities in people's code in a reverse of arxiv.org/abs/2502.174...
July 24, 2025 at 5:38 AM
Timeline where companies comply and their LLMs start putting vulnerabilities in people's code in a reverse of arxiv.org/abs/2502.174...
Reposted by RatFace5000
maybe RLHF is doomed because human revealed preferences are bad
i legitimately cannot see as a matter of design how to go from where we are to non-asskissing LLMs. it's fundamental to how finetuning data is sourced: people rank ass-kissing data higher.
July 20, 2025 at 10:28 PM
maybe RLHF is doomed because human revealed preferences are bad
Reposted by RatFace5000
It’s criminal that this game isn’t funded yet. Avery is the best in the biz at designing games that are unique, compelling, and still intuitive to pick up and play. I can’t wait to get Going for Broke to my table. Check it out and back the campaign ASAP!
The latest update for Going For Broke talks about one of the biggest reasons why this game matters to me. I hope you'll give it a read! Four days remain in the campaign.
www.backerkit.com/c/projects/b...
www.backerkit.com/c/projects/b...
Going For Broke - a sitcom game about making rent. Project Update: Why This Game Matters to Me
I've been designing roleplaying games for nineteen years – the entirety of my adult life. The games that have made it to print have generally been ones that really mean somet...
www.backerkit.com
March 6, 2025 at 10:12 PM
It’s criminal that this game isn’t funded yet. Avery is the best in the biz at designing games that are unique, compelling, and still intuitive to pick up and play. I can’t wait to get Going for Broke to my table. Check it out and back the campaign ASAP!
Reposted by RatFace5000
I simultaneously think that AI (as in LLMs) is probably a net negative socially, and that many, many high profile professional critics of it don't have a fucking clue what they're talking about.
like one of the big things about 404 media, Brian Merchant, Paris Marx, Ed Zitron, etc., is that they neither know nor care how the subject of their criticism actually works
January 23, 2025 at 11:28 PM
I simultaneously think that AI (as in LLMs) is probably a net negative socially, and that many, many high profile professional critics of it don't have a fucking clue what they're talking about.
maybe social networks should have a ranking system I should be able to opt out of seeing ELO 800 posters
i see plenty of factual conversation on here. i just also see a lot of posts like this complaining that everyone else won't cede that their thoroughly discredited ideologies should be taken seriously and argued with from first principles for the 10,000th time
November 25, 2024 at 5:12 AM
maybe social networks should have a ranking system I should be able to opt out of seeing ELO 800 posters
Disgusting that people are laughing at Wiley Coyote being crushed by an anvil after the collapse of the contraption he himself constructed. Basic empathy failure of the left.
October 18, 2024 at 6:15 AM
Disgusting that people are laughing at Wiley Coyote being crushed by an anvil after the collapse of the contraption he himself constructed. Basic empathy failure of the left.
spell borguaise shit I can't even get the right number of s in disposession
October 18, 2024 at 6:12 AM
spell borguaise shit I can't even get the right number of s in disposession