If you can train arbitrarily large models and put your weights online, without being required to do as much as email the government, so they know what you are doing, I would resolve this market positive.
For example, if the AI act looked something like the US executive order, then I would still resolve this market as “Includes Regulation”, because it requires you to report training runs using more than 10^26 flops.
For this market, reaching an agreement is not the same as the EU AI Act entering into force. It means that, according to reputable media sources, some compromise has been agreed upon. Since foundation models are the main point of disagreement in the trilogue, I expect this to be easy to resolve by the end of the year.
I'll use Zvi's house rules on this market.
1,000🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ337 | |
| 2 | Ṁ191 | |
| 3 | Ṁ96 | |
| 4 | Ṁ57 | |
| 5 | Ṁ42 |
People are also trading
Perhaps this is essentially equivalent to this question, but it was defining foundation models through paragraphs that I couldn't find in the linked document: https://manifold.markets/MarcusAbramovitch/will-the-eu-ai-act-implement-regula
Why anyone would not do this:
At a meeting of the Telecom Working Party on Thursday, a technical body of the EU Councils of Ministers, representatives from several member states, most notably France, Germany and Italy, pushed against any type of regulation for foundation models.
Leading the charge against any regulation for foundation models in the AI rulebook is Mistral, a French AI start-up that has thrown the gauntlet to Big Tech. Cedric O, France’s former state secretary for digital, is pushing Mistral’s lobbying efforts, arguing that the AI Act could kill the company.
Meanwhile, Germany is being pressured by its own leading AI company Aleph Alpha, which Euractiv understands has very high-level connections with the German establishment. All these companies fear the EU regulation might put them on a back foot compared to US and Chinese competitors.
Ups. Used a bad title. I should have spent less time defining loopholes and rather define what “reaching an agreement” means. I am actually not that familiar with the technicalities here. I guess what I am actually looking for is whether people will call the negotiations to have come to “some agreement” and to “not be in deadlock anymore” whatever that means. I'll probably resolve according to whether reporters still report a "deadlock "
I am happy to bail out money to people who bet on the “EU AI Act Negotiations Don’t Reach an Agreement in 2023” option, as that seems less of a hassle than making a new question.