←
Matt Perault1,001 posts
Matt Perault avatar

Matt Perault

@MattPerault

Head of AI Policy @a16z. Fmr director of @UNC_TechPolicy and Facebook public policy.

◷ Joined September 2012▧ 58 Photos and videos
Matt Perault avatar
Matt Perault@MattPerault·2 Jun 2025

1/ Are the right AI rules... zero rules? Not for Little Tech. To ensure AI thrives in the long run—and that startups can compete at the frontier—we need smart regulation. Here’s what we believe at @a16z 🧵

Matt Perault avatar
Matt Perault@MattPerault·23h

How can we measure catastrophic AI risk? @Arena’s @ml_angelopoulos challenges how we assign probabilities to those scenarios. He argues for breaking broad risks into specific behaviors and capabilities we can test in order to identify and address the factors that could contribute to harm.

1:23
Anastasios Nikolas Angelopoulos avatar
⟳ Matt Perault retweeted
Anastasios Nikolas Angelopoulos@ml_angelopoulos·Nov 16

It was a great conversation @MattPerault ! Thanks for having me. We need a neutral layer of evaluations that evolves as quickly as the models do.

Even the people building AI evaluations have a hard time keeping up with how quickly the technology changes. That’s a reason for policymakers to leave room for new evaluators and better methods to emerge. More competition can mean better information for everyone making decisions about AI. A useful perspective from @ml_angelopoulos, who has experienced this firsthand building @Arena.

0:58
Matt Perault avatar
Matt Perault@MattPerault·Nov 16

Even the people building AI evaluations have a hard time keeping up with how quickly the technology changes. That’s a reason for policymakers to leave room for new evaluators and better methods to emerge. More competition can mean better information for everyone making decisions about AI. A useful perspective from @ml_angelopoulos, who has experienced this firsthand building @Arena.

0:58
Matt Perault avatar
Matt Perault@MattPerault·Nov 13

How do you ensure that AI developers aren't grading their own homework? @Arena’s @ml_angelopoulos offers a clear standard for independent AI evaluation in our latest conversation.

1:16
Matt Perault avatar
Matt Perault@MattPerault·Nov 11

AI leaderboards help shape our understanding of what models can do. What goes into those rankings? How should we use them? And how can public policy support innovation in AI evaluation, rather than freezing it? Anastasios @ml_angelopoulos joins me to explain @Arena’s approach to evaluating AI performance based on one main metric: how it helps people to get real work done. We cover: • Arena’s beginnings as a Berkeley research project • What independence means in AI evaluation (the “you can’t grade your own homework” problem) • What policymakers and the public can learn from Arena’s new Alignment Index • A different approach for measuring catastrophic AI risk • How demand and competition push model developers to improve • Why policy should keep evaluation open to new competitors and approaches I learned a lot from this episode!

0:44
Matt Perault avatar
Matt Perault@MattPerault·Nov 2

What does it mean to serve clients well as AI changes the practice of law? @Harvey’s chief legal officer John LaBarre on how lawyers’ professional responsibilities could evolve alongside the technology.

0:46
Anne Neuberger avatar
⟳ Matt Perault retweeted
Anne Neuberger@AnneNeuberger·Nov 2

I’m pleased to welcome Andy Baker as an advisor to a16z Global Affairs. I’ve valued Andy’s perspective on national security and look forward to working together to help our founders build partnerships that strengthen America and our allies.

Anne Neuberger @AnneNeuberger Nov 2 Andy Baker, former U.S. Deputy National Security Advisor, joins as Advisor to a16z’s Global Affairs Technology is now central to how countries defend themselves, grow their economies, and provide for their citizens. For American innovators, this creates both an opportunity and a responsibility: to

Matt Perault avatar
Matt Perault@MattPerault·Nov 2

Are coaches too quick to bench players in foul trouble? College basketball tips off this week, and with it comes a familiar ritual: a player picks up a second foul in the first half, and the coach sends him to the bench. Same with a third foul early in the second half, or a fourth at any point. But in many of these games, the benched player never even fouls out. Coaches guarantee their best players lose minutes to avoid the risk that they might lose minutes. It’s coaching orthodoxy, but it costs teams points and wins. In a post out today on my personal blog, I explore protective benching, its costs, and how to fix it.

Post attachment