Pablo Moreno 🔸 🇪🇺 🇺🇦

Personal opinions only, not from my employer

[Marc Thiessen @marcthiessen]

Happy Birthday America!

[Jason Crawford @jasoncrawford]

https://t.co/0q06qO2lB0

[Jaime Sevilla @Jsevillamol]

We want to hire someone obsessed with finding the best people to join Epoch. Come help us grow!

[Pablo Moreno 🔸 🇪🇺 🇺🇦 @Pablomorecasa]

@Noahpinion I think "they abolished slavery first" is the most important contribution to the list.

[Anton Leicht @anton_d_leicht]

the way to actually do this is still the same as in March: trade compute and integration with legacy industry for access and presence.

show you're serious by offering deals and building things, skip the public statements--they just antagonise USG and force Anthropic to deny.

[Anton Leicht @anton_d_leicht]

what does it look for a continent to go 'all in' on something? I think robotics / advanced manufacturing is definitely high on the list of European strategic hopes as they relate to AI, and fusion sounds great to lots of people too (though is ofc dismissed as speculative).

but that doesn't necessarily mobilise the political capital to specifically subsidise these sectors over incumbent industries or energy providers. the energy crisis is happening right now, and spending fungible resources on long-term solutions instead of short-term fixes is hard; heavy industry is facing China disruptions right now, and trying to prop up new niche players over keeping it afloat is also a hard sale.

at any rate, the most effective policy you can probably make for all your ideas is just good supply side reform, innovation ecosystem / capital stuff, etc - which is already substantively overdetermined and just politically tricky. (and I will say no one is actually thinking about competing on training / inf - mostly the idea behind building out inference is just some minimal sovereignty because remote access is so much easier to cut off, and because you might get better access guarantees if they're anchored in infra).

[Alec Stapp @AlecStapp]

It’s hard to find something economists agree on more than “rent control is a bad idea”

And that’s because of empirical research results like this:

[Nuño Sempere (SF until July 9th) @NunoSempere]

It's also just fairly frustrating to see bad news about Africa every week, and now that the amount of state capacity to fix it is just too high for me to do anything about it

[Dwarkesh Patel @dwarkesh_sp]

What does the next training paradigm look like?

0:00:00 – The big research bet the labs are making
0:02:12 – Grindability is just as important as verifiability
0:06:10 – Will RLVR alone generalize?
0:08:41 – Getting the learning back to the weights
0:15:22 – Dreaming
0:17:23 – What 2027 looks like

Also on YouTube, pod feed, and Substack.

[Pablo Moreno 🔸 🇪🇺 🇺🇦 @Pablomorecasa]

@Jsevillamol I'm Epoch-pilled

[Simon Grimm @Simon__Grimm]

Assume European governments release a good AI model that is only slightly worse than what is made by Anthropic or OpenAI.

What happens next?

In a new post, @pietergaricano and I describe the likely outcome: https://t.co/nzPvCmaRMv

The model would provide temporary independence. But a model that is close to the frontier one month will be far behind it six months later. To stay a fixed distance behind the frontier, the model would need constant improving. To fund that improvement, the model would have to make money.

The market for artificial intelligence is not like a market for goods, but more like the market for talent.

In many industries, the best workers make much more than normal workers: the best football players and musicians make thousands of times the industry average; in AI research, the best engineers earn tens of millions of dollars.

Artificial intelligence is similar. Revenues are extremely concentrated in two companies, OpenAI and Anthropic. This is explained by imperfect substitution: new models can do tasks that no combination of old models can do (see GPT 5.6 solving an open Erdős problem). But new models can also do existing tasks up to a standard that weaker models can't, similar to how people can write books or compose music, but no committee could write War and Peace or come up with Bach’s cantatas.

As a result, people and companies often prefer better models. This is even true if a new model is slightly more expensive and only slightly better. Many things in business and life have nonlinear outcomes: a slightly better sales pitch can win a contract; a slightly better job application is the difference between getting and not getting a job. Across such tasks, a slightly better model is worth all of the upside it creates.

We do not think open-source models will be able to meaningfully undercut frontier companies. This is because they are worse at the most economically valuable tasks and will likely only take on easy or unimportant tasks. If cheaper fast-follower models take over such tasks, we might again end up in a world that looks like the market for knowledge work: a lot of work is routine and modestly paid, while the best work on high-value problems and make enormous returns.

Revenue concentration has enormous consequences for how governments should think about investing in frontier AI development.

Such investment is unlikely to be profitable. To justify investment, you might point to the national security benefits of such a model. But then we should be clear-eyed about the fact that it is governments who will pick up most of the tab. If you do a crash effort to build a frontier model, but end up two months behind the frontier, you will have spent hundreds of billions of dollars, without any market actors using your model. The same is true if you try to remain six months behind the frontier. Without financial returns (and without distillation, given that you'd need government funding), staying even a fixed distance behind the best models will be achievable only with exponentially growing subsidies.

Quite a bit of capital is being invested in companies that are developing or serving models behind the frontier: Mistral has raised at an €11.7 billion valuation, Baseten at $13 billion, Fireworks at $4 billion, and AMI Labs at €900 million. It is fine for private investors to take these risks. But European governments and those advising them should not assume there is a great opportunity being left on the table here. Instead, they should make life maximally easy for those private efforts.

[Pablo Moreno 🔸 🇪🇺 🇺🇦 @Pablomorecasa]

A key takeaway for me is that I believe cultivated should be viable and could play a key role in ending factory farming. However, it will by itself not be sufficient: demand-side measures are probably needed.

You play with the model and assumptions here: https://t.co/TAUtoRYp5G

[Pablo Moreno 🔸 🇪🇺 🇺🇦 @Pablomorecasa]

I have written a blog post for the #EffectiveAltruism forum on #CultivatedMeat. I review techno-economic analyses and present a demand-side economic model to predict demand based on key features like taste, price, neophobia...

[Pablo Moreno 🔸 🇪🇺 🇺🇦 @Pablomorecasa]

You can read the blog post at https://t.co/GKmILtW70P

[NXT EU @NXT4EU]

Meloni: the EU being a market is no longer enough, we need Europe to become a genuine power.

[Alex Petropoulos 🤠 @AlexTPet]

Did my first irl Europe2031 presentation yesterday. Was really happy with the turnout and all the engaged and interesting questions :)!

[Pablo Moreno 🔸 🇪🇺 🇺🇦 @Pablomorecasa]

@Noahpinion It's sort of funny because in Spain, I do not get the feeling that people refuse to install AC, quite the contrary, but perhaps it's just my impression.

[Guli Moreno @GuliMoreno]

Hoy, otra startup fundada por un español ha levantado de @a16z en Serie A. $30M, que se dice pronto.

La tercera en año y medio. Se calienta la cosa 🔥🇪🇸

[Lewis Bollard @Lewis_Bollard]

The odds of the Save Our Bacon Act passing have fallen from 50%+ to ~15% on betting markets.

That's thanks to the animal advocates and family farmers who united to oppose it — and to all of you who called your senators and spoke up.

But 15% is not zero. Let's get it there.

[Luis Garicano 🇪🇺🇺🇦 @lugaricano]

It is quite simple. We don't have startups because there are no incentives to start anything up.