No, every startup is NOT raising their first round for $5M at $50 post these days. (but a tiny few are) Data below shows the valuation cap ranges for first rounds on SAFEs over the past 12 months. As round size rises, so does val cap. And deals sign at all sorts of prices

had a blast and know this convo goes against a lot of sales advice folks receive …

Recently, a tutorial/textbook on diffusion models was released on arXiv. Looks like a great resource, starts from the beginning, explaining diffusion models from variational, score, and normalizing flow perspectives, discusses sampling strategies, distillation, fast generation,

I just read a paper co-authored by math legend Terence Tao and researchers at Google DeepMind that completely broke my brain. What if AI's real breakthrough in science isn't just solving problems, but inventing entirely new ways to solve them? A thread on a wild new discovery

Introducing Nested Learning: A new ML paradigm for continual learning that views models as nested optimization problems to enhance long context processing. Our proof-of-concept model, Hope, shows improved performance in language modeling. Learn more: https://t.co/fpdDlYaleL

I'm 24, dropped out of Duke, joined YC, and bet everything on a startup. Here's what actually happens when you go all-in: 1. Your co-founder relationship matters more than your product: You can pivot products. You can't pivot people. Most startups die from founder breakups, not

Forget Individual Genius—The Future of AI Is All About Family Trees (1/10) The secret to building a superintelligent AI isn't to reward genius. It's to reward good parenting. A new paper just dropped that completely changes how we think about AI self-improvement, and it's a

Stanford just published a huge 470-page study 📕 "The Principles of Diffusion Models" Explains how diffusion models turn noise into data and ties their main ideas together. It starts from a forward process that adds noise over time, then learns the exact reverse. The reverse

The "Reasoning Pattern": The One Thing Your LLM Is Missing AI companies are spending millions on a huge, expensive mistake. They're paying armies of people to teach AI how to "think," step-by-step. A new paper reveals there's a way to get the same results for a fraction of the

What percent of Series A startups go on to gain at least 10x in valuation? For Series As from 2017-2020: - 15% are now worth at least 10x - 51% are worth 2x or less (44% never got more valuable at all) Across the whole distribution from 2017-2025, there are 5 companies in this

🔥 MIT just exposed every top AI model and it’s not pretty. They built a new test called WorldTest to see if AI actually understands the world… and the results are brutal. It doesn’t just check how well a model predicts the next frame or maximizes reward it tests whether it

Very happy to see the final outcome of our continuous multi-year efforts on meta-gradient RL published in Nature: state-of-the-art RL algorithms discovered by gradient descent! Please also check out the original meta-gradient RL paper https://t.co/BdnyZVxfyz and its online

🚨 The AI paper everyone’s quietly freaking out about. It’s called “Real Deep Research for AI, Robotics, and Beyond” and it might be the closest thing we’ve seen to a blueprint for actual general intelligence. Here’s the wild part: Instead of memorizing patterns, this system

"The point of venture capital is to allow a founder to run an experiment against a hypothesis that is knowable, testable, where a true dramatically changes the EV of the company. If you look over history that's when it has performed best, and that's really how it works. Using a

Sequoia unveils $950M in new early-stage funds as it strives to be ‘only as good as our next investment’ https://t.co/0Y0h5Z8Bjf via @techcrunch

🚨 This is wild. A new paper from the Ling team just dropped "Every Attention Matters" and it quietly rewrites how long-context reasoning works in LLMs. Their new Ring-linear architecture mixes Softmax and Linear Attention, cutting inference cost by 10x while keeping SOTA

1/16 I just fell down a rabbit hole reading a new paper from economists at MIT & Harvard. Their prediction is wild: We're on the verge of a "Coasean Singularity"—a future where AI agents make markets so efficient that the very idea of a 'company' starts to crumble. 🤯 A thread

Meta's Chief AI Scientist Yann LeCun offers a critical take on the humanoid robot boom. Speaking at MIT, LeCun claimed the "big secret" of the industry is that current companies "have no idea" how to make their robots "smart enough to be generally useful." He argues that while

long post bc I have a feeling that this AI Growth Lab proposal might go under the radar but imo is genuinely one of the biggest things UK might do to thrive in a world of transformative AI and get many more companies like @elevenlabsio @wayve_ai @delian_ai @clovefi & many more

Google just did something nobody saw coming. While Big Tech burns billions on massive cloud infrastructure, they dropped a bombshell that makes all of it optional. It's called Coral NPU, and it's the first chip architecture that runs actual LLMs on devices so small you wear