Tools

Search

Explore

Videos Channels Figures

Atmrix

About

Tools

Search

Explore

Videos Channels Figures

Atmrix

About

BM

Bob McGrew

06/17/25

@ Sequoia Capital

In 2025, reasoning will become the defining focus in AI development, as pre-training reaches diminishing returns.

Video

SC

The Breakthroughs Needed for AGI Have Already Been Made: OpenAI Former Research Head Bob McGrew

@ Sequoia Capital

The Breakthroughs Needed for AGI Have Already Been Made: OpenAI Former Research Head Bob McGrew

06/17/25

Related Takeaways

BM

Bob McGrew

06/17/25

@ Sequoia Capital

The diminishing returns in pre-training mean that significant increases in intelligence will require exponentially more compute resources, making it harder to achieve breakthroughs.

BM

Bob McGrew

02/01/25

@ Y Combinator

The current focus in AI development is on scaling models effectively, as the foundational challenges of reasoning and computation have been addressed.

YC

Y Combinator Cast

03/11/25

@ Y Combinator

The era of scaling pre-training may not be completely over, but reasoning now appears to offer the most potential for maximizing gains from scaling compute.

SA

Sam Altman

05/12/25

@ Sequoia Capital

In 2025, I expect coding to be a dominant category, with AI potentially making significant scientific discoveries.

YC

Y Combinator Cast

03/11/25

@ Y Combinator

Reasoning is expected to be a core capability of future models, and the two paradigms of unsupervised pre-training and specialized reasoning complement each other well.

SC

Sequoia Capital Cast

02/25/25

@ Sequoia Capital

The year 2025 is anticipated to be significant for AI agents, with a focus on their breakout potential in various applications.

MA

Marc Andreessen

12/15/24

@ Chris Williamson

By 2028, I believe we'll have advanced AI systems capable of solving complex problems that humans can handle, thanks to significant ongoing work in AI development.

BM

Bob McGrew

02/01/25

@ Y Combinator

The shift from merely scaling LLMs to incorporating reasoning techniques has opened new avenues for AI development, allowing models to perform better by thinking longer and more deeply.

BM

Bob McGrew

02/01/25

@ Y Combinator

Achieving AGI requires not just pre-training but also the ability to reason and utilize additional computational resources effectively to improve outcomes.