AI Summary
5 min readIn December 2019, Greg Brockman had just finished training GPT-3. As everyone headed out for holiday vacation, he looked at the model "sitting on a shelf" and felt a pang: every day no one explores what a new capability can do is a day lost to the world. He canceled his plans and spent the holidays poking at the model, trying to teach it to sort numbers (it didn't work well). That same impulse, he says, is what the world needs now—only the stakes are far higher.
The AGI Era Is Here, But It's Jagged
When OpenAI co-founder Greg Brockman and Ilya Sutskever estimated AGI timelines back in 2016-2017, their math on compute and Moore's Law pointed to about 15 years, or maybe 10 if they were willing to build massive supercomputers. That timeline has proven broadly accurate, and Brockman now says "we're in the AGI era." But he doesn't mean a single moment of arrival. AGI has turned out to be "less of a point in time and more of this sort of fuzzy spectrum." The Astra model (which the company internally considered worthy of the GPT-6 label because so many research bets came together at once) is what convinced him: Brockman has seen it "run coherently for 24 hours to go accomplish tasks" across a wide variety of domains.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of a16z Show
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:00) **Opening: The AGI Era is Here** - Greg Brockman declares that OpenAI considers the current moment the AGI era, citing Astra's sustained performance.
- 2 (02:09) **Timeline Predictions and the Macro View** - Greg and Ben Horowitz discuss how the 10-15 year AGI timeline predicted in 2016 now feels accurate in retrospect.
- 3 (04:06) **The Bottleneck: Compute, Safety, and Distribution** - Greg argues that the hardest part of the future won't be model capability, but getting it to everyone safely.
- 4 (04:56) **The Evolution of Safety: From Filters to Architecture** - Ben and Greg trace the safety problem from simple content filters to deep architectural alignment for autonomous agents.
- 5 (08:00) **Coordination and the Need for a Safety Community** - Greg discusses why no single company can solve safety alone and why industry-wide coordination is critical.
- 6 (09:27) **The Hugging Face Watershed: Defenders' Window Opens** - Greg explains why the Hugging Face incident is a turning point for cybersecurity and why defenders have a unique, urgent opportunity.
- 7 (11:53) **The 10,000 Agent Math Problem and the Defense Factory** - Ben and Greg discuss the dual-use nature of agent swarms: solving hard math and finding security holes.
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
Ben Horowitz and Erik Torenberg sit down with OpenAI co-founder and President Greg Brockman to discuss why he believes AI has entered a new phase, what OpenAI’s latest models reveal about the path to AGI, and the safety and security challenges that come with increasingly capable systems.
Greg explains why computer use represents such an important step for agents, including models that can work coherently for 24 hours and interact with software through the same interfaces humans use. He also shares how OpenAI deployed 10,000 agents to tackle the Navier-Stokes problem, and why advances in mathematical reasoning could translate into new approaches to science, software, and cybersecurity.
Ben, Erik, and Greg also dig into the “defender’s window” for cybersecurity, how AI could reshape work and entrepreneurship, and what the AI assistant of the future might actually look like: persistent, proactive, personalized, and capable of doing work on your behalf rather than waiting for another prompt.
Resources:
Follow Greg Brockman on X: https://x.com/gdb
Follow Ben Horowitz on X: https://x.com/bhorowitz
Stay Updated:
Find a16z on YouTube: YouTube
Find a16z on X
Find a16z on LinkedIn
Listen to the a16z Show on Spotify
Listen to the a16z Show on Apple Podcasts
Follow our host: https://twitter.com/eriktorenberg
Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.
Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.