Many people mistakenly think that the main battles in society are between the haves and the have-nots. But what’s really happening is that the haves are trying to seize money, resources, and status from the have-mores—and they disguise this motive as concern for the have-nots.
You even saw this recently in New York, where Mamdani failed to win over poor and working-class voters. But he did win support from the upper-middle class. Then, when you move to the very top of the income scale, the dynamic flips again—the wealthiest New Yorkers were more likely to back someone like Cuomo.
So a lot of these political battles, especially around socialism, aren’t being driven by the poor. They’re being driven by people who are doing well, but who are angry they’re not at the very top. It’s less about the 99% versus the 1%, and more about the 9% versus the 1%. The bottom 90% aren’t really part of that fight.
I’m reading We Have Never Been Woke by sociologist Musa al-Gharbi, and he makes a similar point. He notes that during the Occupy Wall Street protests in the early 2010s, most of the participants were white, college-educated, upper-middle-class people—people who were objectively doing fine, but were upset they were in the top 10% instead of the top 1%.
The last few years were about proving what AI can do. Now, it's about building with it.
We've put together a new list of startup ideas we're especially excited about— ideas that treat AI not as a feature, but as a foundation.
https://t.co/AbQWcpKCZX
why is AI
- stuck in chat?
- fixated on one-shotting tasks?
- always booking flights/making dinner plans?
Introducing Dex v0. The world's first AI browser copilot.
The bottleneck is not the model. (our CTO disagrees)
It’s the lack of imagination.
@dexbythirdlayer
- actually does your busywork, just how you like it
- doesn’t require yet another app, it lives in Chrome
- doesn’t take over your screen, it makes you faster
Early preview is live. Comment & I’ll reach out personally.
Startups have to rediscover zero bug bounce
Every few sprints get to zero P2 bugs before you start new feature work. Otherwise your product will be a low quality mess forever.
One part of great craft is getting to zero known bugs regularly.
🚀 New video case study: @YCombinator hit 90% end-to-end test coverage on its mission-critical applications portal in just 14 days with only 2 engineers, powered by Spur's AI agents!
📊 90% coverage in 2 weeks
⏱️ 100+ QA hrs saved per batch
🐞 Critical bug caught the night before peak submissions, preventing an outage
We're excited to share that @spurtest_ has raised $4.5 million in Seed funding, led by @firstround to allow teams to own high quality testing
@sneha8sivakumar & I met during freshman year, we joined @ycombinator last may, & we just launched--as featured on @BusinessInsider 💜
One of the best parts about YC is that the community of great builders are so eager to share their learnings with each other. Thanks @hsinleiwang, @fin465 and @emkara for coming back this week to speak with the S25 founders.
In the words of Nelson Mandela: it always seems impossible until it’s done.
My friends, it is done. And you are the ones who did it.
I am honored to be your Democratic nominee for the Mayor of New York City.
NEW: @JustinBrannan believes “we’ll get there,” on raising taxes for the wealthy to pay for some of @ZohranKMamdani plans.
“We need help from Albany and I think the majority of Americans support raising taxes on the rich,” says Brannan
More tonight on PIX on Politics
One week ago today, we shocked the establishment and redrew the political map of New York City with a campaign relentlessly focused on the needs of working people.
Let's break it down.
There's so much low-hanging fruit with current AI. Tons of simple jobs that could already be automated if you had good data access and company context.
But what's really exciting for superintelligence is that it's time for research again.
For over a decade, we've known the ingredients: large neural nets, lots of data, GPUs, end-to-end training. We scaled data and its quality, compute, and model size. It worked better and better until we crossed thresholds and got these emergent intelligent properties.
Now we're past the pure scaling phase and it's time to ask the deeper questions: What is true intelligence? How do we get to superintelligence? What does exponential intelligence growth actually mean?
I think there are different dimensions of intelligence with different upper bounds. Some are astronomically far away. Others aren't that hard.
But the AI researcher in me that still beneath the entrepreneur is pretty excited about this.
Proud to announce our multi-year collaboration with OpenAI as a global services partner.
Together, we’ll deliver transformative, scalable AI solutions—combining Open AI’s AI models with Globant’s digital engineering and AI delivery frameworks.
Learn more here: https://t.co/Ivco8d5rEo
We are excited to announce that @shengjia_zhao will be the Chief Scientist of Meta Superintelligence Labs!
Shengjia is a brilliant scientist who most recently pioneered a new scaling paradigm in his research. He will lead our scientific direction for our team.
Let's go 🚀
I’m excited to be the Chief AI Officer of @Meta, working alongside @natfriedman, and thrilled to be accompanied by an incredible group of people joining on the same day.
Towards superintelligence 🚀