/Tech1d ago

Demis Hassabis rejected Facebook's higher acquisition bid after Mark Zuckerberg failed a private AI strategy test

Story Overview

During DeepMind's 2013 acquisition talks, Demis Hassabis visited Mark Zuckerberg's Palo Alto home and steered the conversation from AI's potential toward VR, AR, and 3D printing to probe priorities. Equal excitement for every technology convinced Hassabis that Facebook did not see AI as singularly transformative, so he accepted Google's lower offer instead.

594.4K136960549.3K

#756

Original post

MTS@MTSlive

SITUATION EXPLAINED: Demis Hassabis secretly tested Mark Zuckerberg at dinner before deciding not to sell DeepMind to Facebook.

• Facebook offered more money for DeepMind than Google • At Zuckerberg's home, Hassabis brought up other hot tech of the era: VR, AR, 3D printing • Zuckerberg sounded equally excited about all of them • That was enough for Demis to make his decision

Demis Hassabis: "That told me what I needed to know. Facebook offered more money, but I wanted somebody who really understood why AI would be bigger than all these other things."

@theojaffee: "This is a problem I have with LinkedIn. They're excited about technology only very abstractly, and they don't understand why AI is not like the other technologies."

Henry Shevlin@dioscuri

*taps sign*

12:47 PM · Jun 23, 2026 · 21.3K Views

Strategy Signal

The dinner test exposed mismatched convictions

Hassabis treated the evening as an informal benchmark of strategic focus rather than a sales pitch. Zuckerberg's uniform enthusiasm across disparate trends signaled that AI would compete for attention inside Facebook rather than define its direction.

Acquisition Insight

Larry Page's stance tipped the scale

Google's founder conveyed a clearer sense that AI held outsized importance, aligning with Hassabis's own assessment of the technology's trajectory. The book excerpt leaves the exact bid gap and any follow-up negotiations unspecified.

Sentiment

Positive users praise Hassabis for recognizing AI as a civilizational inflection point instead of a routine investment, while negative users dismiss the story as dumb and insult those involved.

Pos

57.5%

Neg

42.5%

21 comments with sentiment.

Cluster Engagement

Digg Deeper

No Digg Deeper questions have been answered for this story yet.

Posts from X

Most Activity

VIEWS5.9KBOOKMARKS5LIKES73

Zy@ZyMazza

@dioscuri You could immediately rank the relative importance by the safteyism around them.

VR-none AR-moderate privacy concerns 3D printing-a smattering of legislative proposals around firearms AI-🚫⛔⏸️🙂‍↔️😤

1d5.9K735

RETWEETS114

Henry Shevlin@dioscuri

*taps sign*

roon@tszzl

a lot of these people think quantum and ai are roughly the same thing on the same order of importance. just shooting darts in scifi semantic space

1d602.4K4.9K1.1K

REPLIES3

Robert Garritty@GarrittyOf

@dioscuri This is dumb. All sorts of exciting technologies out there. AI is exciting precisely because of the other things it unlocks.

1d5.1K511

Catch Vartan@catchvartan

@dioscuri Why is every anecdote about Hassabis about him testing people?

1d2.9K222

Slavic Muse@SalvicMuse

@dioscuri Midwit thing to make such a decision on lol Could be a million reasons why he reacted like that

1d2.1K152

Conspiracy Apologist@Conspologist

@dioscuri

1d51471

Kart ographien@kartographien

@SalvicMuse @dioscuri In hindsight Demis was obviously correct. Zuck wasted billions on VR which went nowhere, and missed the boat on AI.

1d13010

Andrew Eiche@buddingmonkey

@ZyMazza @dioscuri I wish that were true for VR. There are dozens upon dozens of changes we’ve had to make to our games because the lawyers were “concerned”.

There’s a reason why Meta has a hard limit of 10 years old for the headsets and why Sony puts all sorts of warnings on content.

1d1214

Sonnen@Sonnentorte

@EarthlingRohit @dioscuri Nah. Mark did what he always does. He thought he can fool Demis. And Demis went the … ewww fuck u , Mark way.

1d4411

Paul Torte@PaulTorte

@dioscuri guy working in ai related company thinks ai is very important yeah i believe you XDD

1d1.3K11

BSD@bsd_official1

@dioscuri Hilarious argument 😂. BREAKING: Man in charge of one of the world’s largest tech firms is excited about tech. You should WANT that kind of energy not run from it. Foolish.

1d1.2K9

Michael Mullen@zapadooo

@GarrittyOf @dioscuri You think he would have gotten a Nobel at Meta? I doubt it

1d3515

Rosalia@latereigns

@dioscuri his staff legionnaires literally think a DAG is the peak of SWE, meta is NGMI

1d3.7K2

efe@extliqprovider

@dioscuri lol sundar really understands ai thats why he gave all that compute to anthropic gdm is starving...

1d3K2

oitixion@oitixion

@dioscuri @lu_sichu So because AI will be bigger than these things, we can’t be excited about all of them?

1d9217

Pranav@rajmane_pranav

@dioscuri FYI, this was five years before GPT-1. You cant blame him.

1d2521

Deepak Dhingra@WuTangAI

@GarrittyOf @dioscuri 3d printer exciting and we want to summon a demon that might extinct human race exciting are two different kind of excitements.

1d2174

Robert Garritty@GarrittyOf

@zapadooo @dioscuri Probably not, and Alphabet was definitely a better home than Meta for all sorts of reasons. But Larry and Sergey are even less focused than Mark in terms of their excitement for new technology (Kitty Hawk?).

1d2977

lo@868efx

@latereigns @dioscuri Read Build Systems a la Carte

1d791

Edon Derguti@edon_d

@dioscuri LOL it probably was money issue and not this crap. Demis Hassabis should have never sold. Today Deepmind would have been maybe as big as google.

1d2.3K4