Tools

Search

Explore

Videos Channels Figures

Atmrix

About

Tools

Search

Explore

Videos Channels Figures

Atmrix

About

DH

Demis Hassabis

05/24/25

@ Hard Fork

Our flash models are our most efficient and popular models, and we rely heavily on them internally.

Video

HF

Google DeepMind C.E.O. Demis Hassabis on Living in an A.I. Future | EP 137

@ Hard Fork

Google DeepMind C.E.O. Demis Hassabis on Living in an A.I. Future | EP 137

05/24/25

Related Takeaways

MB

Marc Benioff

12/10/24

@ 20VC with Harry Stebbings

Salesforce builds many models and has a phenomenal AI team, but we also leverage other models and regularly invest in various model companies.

SB

Sergey Brin

05/24/25

@ Google for Developers

The recent introduction of the 2.5 Flash model, which ranks highly on performance leaderboards, demonstrates Google's ongoing commitment to AI innovation and development.

ZX

Zheng Xue

05/30/25

@ LangChain

We utilize proprietary models and APIs to generate insights, allowing us to scale our services and provide timely information to clients during meetings.

MK

Mike Krieger

05/09/25

@ Sequoia Capital

I believe that having our models out in the wild has provided valuable market feedback, which has shaped our product development significantly.

DH

Demis Hassabis

05/26/25

@ Hard Fork

You need big models to train smaller models effectively, and while we have efficient smaller models, they often rely on insights from larger models.

DP

Dwarkesh Patel Cast

05/23/25

@ Dwarkesh Patel

We often underutilize models, expecting them to solve tasks quickly without giving them enough time to iterate and improve.

WA

Waseem Alshikh

03/27/25

@ LangChain

We have developed an internal evaluation system to measure our models, as relying on external benchmarks often leads to data contamination and unreliable results.

MZ

Mark Zuckerberg

04/30/25

@ Dwarkesh Patel

As AI becomes more integrated into various applications, we need to optimize for specific internal use cases, which may lead us to use other models if they better serve those needs.

a

a16z Cast

05/29/25

@ a16z

Our platform not only evaluates models but also helps partners select the best-performing models based on user feedback.