Sabrina Halper Show

Sabrina Halper sits down with Ali Ansari, the 25-year-old solo founder and CEO of Micro1, which just passed $1B in ARR. But it all began when Ali moved from Iran to the U.S. at age nine, going from flipping textbooks to building a company that supplies the human expertise used to train AI models.
We get into Micro1’s pivot from AI recruiting to training data, growing to $1B ARR in 2.5 years, whether experts are training their replacements, how to teach AI subjective judgment, why robotics is a data problem, the stakes of selling training data to China, and whether the AI economy can justify the infrastructure being built.
Ali Ansari: https://x.com/aliansarinik
Sabrina Halper: https://x.com/SabrinaHalper

SPONSORS
Presented by micro1: https://micro1.ai They're building the infrastructure for advancing intelligence through expert human data, real-world training environments, and contextual evaluations.

TIMESTAMPS
00:00 Intro
02:56 Growing up in Iran and moving to America
07:37 From eBay to textbook flipping
13:54 Building micro1 and the pivot to training data
19:39 Why internet data is the worst data
22:41 Are experts training their replacements?
29:27 Creativity, alignment, and subjective judgment
36:01 The robotics data bottleneck
45:11 Ordinary vs. Extraordinary Training Data
43:56 America selling data to China
53:56 $1B in 2.5 years
56:53 Can the AI economy deliver?

Yinon Costica co-founded Wiz, which Google acquired for $32 billion in the largest acquisition in its history. I sat down with him at Wiz HQ in Tel Aviv to talk about the race to secure a world increasingly run by AI.Wiz on

X: https://x.com/wiz_ioSabrina Halper on X: https://x.com/SabrinaHalper

PRESENTING SPONSOR
The Sabrina Halper Show is presented by micro1.micro1 is building the infrastructure for advancing intelligence through expert human data, real-world training environments, and contextual evaluations.
Learn more: https://www.micro1.ai/

TIMESTAMPS
00:00 Intro
01:15 Yinon Costica, co-founder of Wiz
02:00 IDF's Unit 8200, and meeting his co-founders
06:30 Leaving Microsoft to start again
08:30 The customer conversations that created Wiz
11:30 From $1M to $100M ARR in 18 months
14:15 How to scale without breaking
17:15 Wiz Atlas: automating vulnerability research
22:45 Can AI give defenders the advantage?
25:15 Cybersecurity's AI twilight zone
27:30 Should you give AI agents your credit card?
32:00 Trust, Big Tech, and the software supply chain
34:45 Open-weight models and security
36:30 Vulnerabilities found at DeepSeek
38:30 Preparing for AI-powered attacks
43:30 What Google unlocks for Wiz
44:30 How Wiz made the shift to AI
47:15 Cybersecurity's biggest unsolved problem
50:00 Securing agents with access to sensitive data
52:15 What AI hacking agents can already do
54:00 An internet built for agents
56:15 The human role in cybersecurity
58:30 Is it too early to pick AI's winners?
59:45 Why the future could be more secure
Subscribe for more conversations with the founders, scientists, and researchers shaping what comes next.#Wiz #AI #Cybersecurity

Alex Atallah co-founded OpenRouter in 2023 and recently signed an agreement to be acquired by Stripe for a reported ~$7B+. Before that, he co-founded OpenSea, the world’s largest NFT marketplace. Alex is a deep thinker and a one-of-a-kind company builder. I was honored to sit down with him again.We also talk about why OpenRouter and Stripe fit together, and their shared belief in an economy where value can come from many different models, businesses, and agents, rather than a handful of giant labs. We get into how Alex finds ideas before they seem obvious, what OpenRouter has taught him about how people actually use AI, why there won’t be one model that wins everything, and how agents will change the way we use software, spend money, and interact with the internet. More than anything, this is a conversation about how Alex thinks, builds companies, and sees the next few years of AI unfolding.Alex Atallah on X: https://x.com/alexatallahSabrina Halper on X: https://x.com/SabrinaHalper XPRESENTING SPONSORThe Sabrina Halper Show is presented by micro1.Micro1 is building the infrastructure for advancing intelligence through expert human data, real-world training environments, and contextual evaluations.Learn more: https://www.micro1.ai/ TIMESTAMPS00:00 Preview01:05 What is OpenRouter?04:17 Why ChatGPT ended his sabbatical07:30 Why Stripe is buying OpenRouter11:26 What happened on Jan 1, 2026?15:22 OpenRouter’s moat18:03 Homepage for all models19:30 Agent purchasing power24:01 Ox Alpha, Stealth models, & brand bias26:50 How AI models get priced & chosen29:29 Chinese open-weight models, evals & security32:39 An internet built for agents34:11 Stablecoins & the economics of AI agents37:25 What humans will still want to own39:21 Why engineers are becoming scientists41:47 What is the next "open" opportunity? 45:55 Our personal relationships with AI48:53 AI advice, accountability & safety51:24 Comparing AI model personalities53:29 Why Alex is optimistic about human creativity56:51 ClosingSubscribe for more conversations with the founders, scientists, and researchers shaping what comes next.#OpenRouter #AI #Stripe

Most of what you hear about AI's impact is a projection. David Holtz has the data.
He’s behind some of the most rigorous studies of AI’s actual impact today. And the results are surprising. His work cuts through the noise around AI to a key question: what is AI actually doing to people, companies, and the economy today?
David is a professor at Columbia, currently on leave to lead AGI economics research at OpenAI. We recorded this before he started there.
If you enjoy this one, subscribe! It genuinely helps the show grow and improve every week.
-
Presented by micro1: ⁠micro1.ai
LinkedIn: ⁠micro1
-
Sabrina Halper
X: https://x.com/SabrinaHalper
Instagram: https://www.instagram.com/sabloves/?h...
Apple Podcasts: ⁠https://podcasts.apple.com/us/podcast...
-
David HoltzX: ⁠https://x.com/daveholtz
-
TIMESTAMPS:
00:00 Intro
01:26 Is AI as significant to the economy as we hear?
04:05 ChatGPT as a general purpose technology shift
06:41 Inside OpenAI: The Shift From Chatbots to Agents
08:56 Why AI Could Make Us Work More, Not Less
10:56 Structuring Companies In a New Way
15:02 What Will the Agent Economy Look Like?22:03 What Becomes a Moat When Everyone Has the Same AI?
25:02 Model Improvement Study: 50% of AI’s Gains Come From the Human
33:03 What Should We Still Learn in the Age of AI?
36:58 The Kenya Experiment: Does AI Democratize Expertise?
45:11 Could AI Let Developing Economies Leapfrog the Internet?
49:50 One-Person Companies, Job Loss & the Future of Work

Matan Grinberg left a string theory PhD at Berkeley to start Factory, betting on full autonomous software engineering in 2023.
It started with a cold email to Sequoia’s Shaun Maguire and a three-hour walk that helped convince Matan to leave his PhD and build a company. Three years later, Factory is valued at $1.5B, with recent reports putting its newest valuation north of $3.5B.
We get into: what string theory teaches you about making decisions, the three phases every company goes through with AI adoption, how to measure ROI of AI spend, why he'd take Chinese open models over a closed-model monopoly, the sleeper-agent risk hiding in open weights, what he thinks about recent security incidents from large labs, and why he believes we're already living in a post-AGI world.
-
Sabrina Halper Show is presented by Micro1. Micro1 is building the infrastructure for advancing intelligence through expert human data, real-world training environments, and contextual evaluations. Learn more: https://www.micro1.ai/.
-
TIMESTAMPS:
(00:00) Intro
(02:00 Young Matan
(5:30) String Theory
(11:10) Leaving academia, the IMDb 250, and getting nerd-sniped by code generation
(16:53) Cold-emailing Shaun Maguire and founding Factory
(18:27) The desert years, and what finally made autonomous engineering work
(20:55) The new bottleneck is deciding what to do
(23:30) The three phases of enterprise AI adoption, and the end of token maxing
(28:18) Chinese open models, and the monopoly to end all monopolies
(30:30) Sleeper agents, poisoned code, and the labs cybersecurity incidents
(34:30) The White House open-weights policy and the distillation problem
(37:30) The coming billion-dollar agent incident
(42:30) "We're already post-AGI" and AI taking on math
(49:30) Building your own software
(52:20) Reflections & best advisors