This resonates.
I’d add that organizations don’t accumulate learning automatically—they accumulate the outputs of their social networks.
Learning begins with people asking questions, sharing expertise, challenging assumptions, and solving problems together. AI can capture, scale, and compound those interactions, but it can’t create them.
Which suggests an interesting implication:
The organizations with the strongest social capital won’t just collaborate better—they’ll build better AI.
The trust boundary Satya describes isn’t only around models or data. It’s also around the human networks that generate the learning those systems depend upon.
Social capital may become one of the most underappreciated sources of competitive advantage in the AI era.
The EU AI Act seems designed to allow AI only for routine tasks while hindering its use in high-level problem-solving.
This will endanger European AI startups and significantly damage EU productivity.
THREAD on our post today in Silicon Continent.
1/9
A wonderful day @HarrisPolicy, celebrating our colleague, Jim Robinson, who has won the Nobel Prize, along with @DAcemogluMIT and @baselinescene, for field-defining work on the political and institutional foundations of peace and prosperity.
@cblatts It can apply a book or other body of knowledge to a movie or other piece of media as if it has watched the movie multiple times. I assume it’s merely trained on data including countless reviews, blog posts from actors, writers, etc., and perhaps the script/screenplay.
The totalitarianism we warned of is happening.
The European Union is at this moment forcing big tech companies to secretly engage in mass censorship. Google and Facebook are, apparently, going along with it.
Only Elon Musk’s X, among the major platforms, is resisting.
A few hours ago, Musk said, “The European Commission offered 𝕏 an illegal secret deal: if we quietly censored speech without telling anyone, they would not fine us. The other platforms accepted that deal. 𝕏 did not.”
And so now, the EU is preparing to punish X with massive fines — up to 6% of total global revenue.
I can't imagine a more egregious form of foreign interference in our domestic affairs than foreign governments demanding mass secret censorship for ideological and political purposes.
Most terrifying is the apparently direct involvement of government intelligence and security agencies, and their intermediaries, in demanding censorship.
This is all happening at the very moment that the same governments, including our own, warn of Russian disinformation through spam accounts that hardly anybody ever sees.
Don’t be fooled by what is happening. Governments and former intelligence officials in Europe, Australia, Israel, Brazil, and Ukraine and other nations are not only demanding censorship but also often spreading their own disinformation.
For example, the EU claims that there is more false information on X than on other platforms, and more censorship would reduce it.
This is false. Only X has the open-source Community Notes system, which allows for real-time corrections to false information. A recent study in JAMA, the Journal of American Medicine, found that 97.5% of the Community Notes were entirely accurate, 2%, partially accurate, and 0.5%, inaccurate.
It’s not perfect. I myself sometimes disagree with them. And of course, science is never standing still, so what appears accurate today might not be accurate in the future.
However, X’s Community Notes uses more speech, not censorship, to provide context to controversial content. You can still agree or disagree with the content, even with a Community Note attached.
What the EU wants is for its committees of experts, not Community Notes, to secretly decide what we can read and say online. This is unethical and unconstitutional.
Another key part of the EU’s disinformation is that “researchers” should have access to X’s internal data, which Musk cut off when he bought Twitter. But those people who want the data aren’t researchers. They’re censorship activists, many of whom have deep relationships with governments in general and intelligence agencies in particular.
If the EU succeeds in censoring X, Facebook, Google, and every other major Internet platform, then there is no free speech. There is only government-controlled speech.
Many people rightly worry about the implications of a single man, Elon Musk, being all that stands between us and foreign governments’ totalitarian censorship plans. I worry about that, too. Our speech is inalienable. It is not something governments give to us.
We need to fight back. While we should be grateful to Musk for standing up to the totalitarians in Europe, Brazil, and Australia, we must build a citizen’s movement to fight back.
There’s three things you can do. First, share this video to raise the alarm. Second, add your email to our list here, so we can involve you in the future. And, third, consider making a donation to build the free speech movement and urge Congress and other social media companies to stand up for free speech against foreign interference.
JULIAN ASSANGE IS FREE
Julian Assange is free. He left Belmarsh maximum security prison on the morning of 24 June, after having spent 1901 days there. He was granted bail by the High Court in London and was released at Stansted airport during the afternoon, where he boarded a plane and departed the UK.
This is the result of a global campaign that spanned grass-roots organisers, press freedom campaigners, legislators and leaders from across the political spectrum, all the way to the United Nations. This created the space for a long period of negotiations with the US Department of Justice, leading to a deal that has not yet been formally finalised. We will provide more information as soon as possible.
After more than five years in a 2x3 metre cell, isolated 23 hours a day, he will soon reunite with his wife Stella Assange, and their children, who have only known their father from behind bars.
WikiLeaks published groundbreaking stories of government corruption and human rights abuses, holding the powerful accountable for their actions. As editor-in-chief, Julian paid severely for these principles, and for the people's right to know.
As he returns to Australia, we thank all who stood by us, fought for us, and remained utterly committed in the fight for his freedom.
Julian's freedom is our freedom.
[More details to follow]
Harvard administration tells alumni they are adopting Chicago-style institutional neutrality principles and won't make "official statements about public matters that do not directly affect the university's core function."
The most clearest and crisp explanation, I've ever heard, of how large language models compress and capture a "world-model" in their weights simply by learning to predict the next word accurately.
Furthermore, how the raw power of these base models can then be tamed by teaching them to follow instructions from humans.
Source: https://t.co/8mQmaIVet3
If you thought the drama here on X was amazing this weekend, just wait 24 hours until all of the best posts get screenshotted and put on Reddit and Instagram.
Then wait another 2 weeks until the content makes its way to LinkedIn.
The accountants and consultants are going to go wild over there.
There is only one obvious winner of the last 100 hours, and it's business school professors who have been looking for a sexy anecdote to get students interested in the principle that corporate structure is really important
One of our key sources of human data is no longer fully “human"!
We estimate that 33-46% of crowd workers on MTurk used large language models (LLMs) in a text production task - which may increase as ChatGPT and the like become more popular and powerful.
https://t.co/SJfKjDM6gX