Perrin Institution @PerrinInst

@Polymarket Not a chance in hell this happens

@ZackKorman I understand the spirit of this statement but it’s implicit that they mean the further development of AI, meaning research without enough benchmarking/safety evals!

@GWR All for it

@vikktorrrre This isn’t a bad take!

Again, while this remains a possibility current models are quite limited in their capacity to be malicious at scale. It is incredibly unlikely that there will be any disaster relating to self-replicating models in the very near future (3-4 months) Because timelines are compressed in this space, there's no telling beyond that point.

@aguerrieri @rynorhn Good idea Alejandro

🚨 MASSIVE: Palantir CEO Alex Karp says the same AI capabilities that help America win wars can easily predict Americans’ behavior and even their thoughts. Calling for stricter limits at home to protect Americans’ right to privacy, he says: “You could comply with the letter of the law and totally eviscerate the spirit of the law here.” Karp says increasingly powerful predictions could extend to “the actual thoughts they’re going to have in a domestic context.” “That’s something we have to really look at and control.” But he argues restricting those same capabilities in warfare could leave America unable to compete against adversaries using AI to predict its actions.

@FT We should be more worried about bad actors using agents in a way adjacent to this.

@birdabo Always was.

@SebJohnsonUK This is actually kind of hilarious

@_XMaster96 @shanebarakat Well somewhat of a serious question. Depends on the thing you’re building but honestly really good prompts and plugins can do wonders sometimes. Skimming is kind of just to ensure it did exactly what you ask

@_XMaster96 @shanebarakat You mean skim right? I barely read code anymore...

@AnotherTechFeed @TechCrunch It’s not over the mark

BREAKING: the UK government publishes official rules on how to use AI: 1. Ask first if you really need to use AI 2. Check if a spreadsheet can do the job before using AI 3. Choose the worst model possible so it uses less energy 4. Keep prompts short to reduce the environmental impact 5. In general, use AI only when necessary, as a climate-saving measure With a mindset like this, we should accept the UK back into the European Union

@AstasiaMyers This is exactly what is happening

perrininstitution.org/careers

@Prathkum This was super insightful, thank you!