This is great, I've read multiple books and watched videos about the attention mechanism. Now that I understand it, this is the clearest example I've seen on how attention works.
"The 2026 FIFA World Cup was the hottest World Cup since the competition began in 1930. A Guardian analysis found that close to 20 percent of matches were played under heat and humidity conditions that FIFPRO, the players' union, has flagged as dangerous enough to warrant delaying or postponing play. Some 20 games were played at wet bulb globe temperatures that hit or exceeded 28 degrees Celsius (82 Fahrenheit)." - Katharine Hayhoe's newsletter
I agree that something must be done, but the hydration break is the wrong solution for two reasons.
First, plenty of countries have been regularly playing under these conditions and will keep doing so after the World Cup. Too hot to play in the afternoon? Play at night.
Second, and more important, it changes the game in a major way because playing for 90 minutes with a single break is not the same as playing with three. I imagine data will say whether this has an effect or not, but I cannot imagine that an endurance player would feel the same if they have to always stop halfway.
Thing is this is a technical justification for something that was done purely to add an extra add break in - there is no point in having the technical discussion.
The problem I had with this World Cup was the over commercialization of it. The ticket prices, the cost of getting to the stadiums, the fucking trophy being presented in some bullshit LV case (like it mattered).
From my perspective I let it slide thinking it was the US’s take on how they do sports so fair enough. I can see now following the hyper commercial US World Cup that FIFAs has dollar signs spinning in their eyes and want to continue this way of working.
Unfortunately it doesn’t align with the values of football. It smacks of the Euro Super League (which would have disrupted the whole pyramid of football). It should be nipped in the bud, Infantino fired, and FIFA continuing as a not for profit.
> We request that the U.S. government support an international effort to develop the technical and governance tools needed to deliberately pace the frontier of automated AI development.
This is quite vague to me, what exactly do you mean?
It means that they request the structures needed to navigate the event horizon of recursive self improvement be put into place. Probably: International agreements, verifiability of an AI pause on the level of graphic cards as well as through physical inspections of datacenters. Like it's done with nuclear energy.
'The Fear' - is not some magic intelligence will come alive and be Skynet, these people are ridiculous.
People will use the compute power to do nasty things, we'll be overwhelmed with bots, regular companies will use them for dark patterns, political actors will use them to cheat.
It will be 100% humans using it for bad stuff, and in not very novel ways.
I generally disagree with the claim that distilling a model is equivalent to training on freely available internet content – mostly due to investment required to turn it into a model – but piracy is another story. Pretty inexcusable.
If they paid for tokens say via subscriptions. Wouldn't that just be same as acquiring books and using them as "fair use" to train? I really see no difference.
They did not need to destroy the models they got with pirated content. Would have been more fine if they would have needed to buy the books afterwards and train based on then, again.
> I generally disagree with the claim that distilling a model is equivalent to training on freely available internet content – mostly due to investment required to turn it into a model
The frontier labs all give free access to their models. Why does “investment” change anything? Anyone who’s ever produced any content, free or otherwise, has invested in doing so.
The only plausible issue I see is that if distillation is being done by creating many free accounts to work around limits on free accounts, that’s a bit… impolite? But if they really wanted to avoid that, they could eliminate free accounts, and require users to sign a real contract governing what they can do with the model.
Of course they don’t want to do that, so they’re stuck in the same world as the rest of us, and they don’t have any real basis to complain about it without being hypocritical.
Nice site, but most of the top biographies are US figures: Edgar Hoover, Frederick Douglass, Lyndon Johnson etc.
I'd like to see more non-US biographies.
This looks great, it looks like something I'd use long term instead of Typescript. I do see a couple of things missing before I could port my web app over.
Integration with -
Design systems such as Shadcn or Radix UI.
An example on how to integrate with an OpenAPI endpoint.
All on the roadmap :) topcoat-ui is already usable with Shadcn components (recent Pr: https://github.com/tokio-rs/topcoat/pull/118) we are still working on pulling in more components.
Re: OpenAPi you want to consume or provide an OpenAPI endpoint? I’m assuming you want to provide one. This is probably on the medium term roadmap after other stuff like UI, tighter ORM integration, email, etc…
This would be consuming an openapi endpoint from my rust service generated using utoipa. Currently I generate an openapi.yaml file and use openapi-typescript to generate the api bindings.
But do they at any useful scale? China is probably the only other country deploying AI at any appreciable fraction of their economy, and it's certainly much less than the US.
reply