🇪🇺🚨 BREAKING: EU Ministers have unanimously adopted the EU hashtag
#AIAct
at today's Council meeting.
The Act will now be published in the official journal of the European Union, starting the clock for compliance.
The European Union has just agreed on a Compromise Text for its EU AI Act.
The EU AI Act will set the global standard for how AI systems must be built. Think of it as the GDPR for AI.
I'm going through the text to highlight important changes... 🧵 1/n
🇨🇭Hopping on the "make it more" trend for ChatGPT, I asked it to make an image encapsulating Switzerland and the kept asking it to make it "more swiss"
Initial image
Keynes famously said that:
"The market can remain irrational longer than you can remain solvent."
I am increasingly convinced this is also true for the market of ideas.
@tracyalloway
@jonathanmaze
For those who don't know, Domino's and Google had their IPOs within a few days of each other back in 2004. And the stocks have performed virtually identical since then.
@clairlemon
Misallocation of capital means people have to accept huge losses. That tends to take time. See the huge deleveraging after 2008
Disruption of supply leads to short term disruption. If the supply disruption ends, demand can return or even surge to fill in what was missed.
The hand strikes again🖖: these photos allegedly shot at a French protest rally yesterday look almost real - if it weren't for the officer's six-fingered glove
#disinformation
#AI
Science funding agency: Please review this proposal.
Me: would be happy to, this is in my area of expertise. My consulting rate is X/hour.
SFA: pay you? Don’t you know we are a prestigious SFA? Thousands of scientists work for us for freeeee. Do your part.
Me: my rate is X/hour.
Annex VIII Section B: Builders of Foundation Models are going to have to disclose details about how these models were trained.
Going to be interesting to see what happens when the EU forces OpenAI/Google/Anthropic etc to open the hood... 3/n
The new Article 68c adds a right to explanation for decisions from high risk AI systems.
This means all AI systems need to have interpretability built in as a fundamental component. 2/n
I'm sure there are people who have thought about the impact of AI on the justice system. If any image, audio or video recording can be created by AI in such a way that it can't be reliably told apart from the real thing, how do we handle such evidence in court?