The Pipeline: Social Media → AI

Same Day · PCAST

On the same day as the verdict, President Trump announced his AI advisory council: Zuckerberg (Meta), Brin (Google), Andreessen (a16z), Huang (NVIDIA). Both events occurred March 25, 2026.

Fidji Simo

Meta 2011–21 OpenAI

Then

VP & Head of Facebook App

Built autoplay video, behavioral News Feed monetization, Facebook Live & Watch. Led the ad-targeting systems central to this week's addiction verdict.

Now

CEO of Applications, OpenAI

Oversees all commercial operations for 800M weekly ChatGPT users — product, engineering, business, finance. Reports to Sam Altman.

She built the autoplay and behavioral video mechanics the jury just examined. She now runs the commercial operations for the AI platform with the most intimate archive of human thought ever assembled.

Kevin Weil

Meta 2016–21 OpenAI

Then

SVP Product, Instagram & Libra

Built Instagram's growth engine. Co-led Libra, Facebook's failed cryptocurrency. Previously SVP Product at Twitter.

Now

Chief Product Officer, OpenAI

Reports directly to Fidji Simo. Oversees all product development across ChatGPT and OpenAI's full product portfolio.

Built Instagram's growth engine → CPO at OpenAI. The two former Meta executives now control the product and commercial direction of the world's most widely used AI platform.

Bret Taylor

Meta / Facebook OpenAI

Then

CTO, Facebook

Co-created the Like button. Built Facebook's platform and developer ecosystem. Left to be co-CEO of Salesforce.

Now

Board Chair, OpenAI

Chairs OpenAI's board while running Sierra AI — an enterprise AI company that uses OpenAI's API. Dual commercial interests.

Facebook CTO → chairs OpenAI's board while his own AI company (Sierra) runs on OpenAI's infrastructure. He governs the platform he profits from.

Adam D'Angelo

Facebook 2008–10 OpenAI

Then

CTO, Facebook (2008–2010)

One of Facebook's earliest CTOs. Left to co-found Quora. Joined OpenAI's board in 2023.

Now

Board Member, OpenAI

The only board member who survived the November 2023 coup that briefly fired Sam Altman — he voted to fire Altman, then reversed course.

Facebook CTO → the sole board survivor of OpenAI's governance crisis. His continuity is the hinge on which OpenAI's board pivoted.

Marc Andreessen

Meta Board 2008– PCAST a16z AI

Then

a16z co-founder — Meta board since 2008

Co-founded Netscape. Co-founded Andreessen Horowitz (a16z). Joined Meta's board in 2008 and has never left.

Now

Meta Board + PCAST AI Adviser + a16z AI investments

On Meta's board, named to Trump's PCAST AI advisory council, and through a16z: invested in SSI, Thinking Machines, OpenAI. Campaigns against AI safety regulation.

On Meta's board for 16 years while funding Meta's AI competitors, and now co-writing US AI regulation. Maximum structural conflict of interest — in every direction simultaneously.

Joel Kaplan

Bush White House Still at Meta

Then

Bush White House OMB → Meta VP Global Policy (2011)

Harvard Law. Bush White House Office of Management and Budget. Joined Meta in 2011 as VP Global Public Policy. Promoted to Chief Global Affairs Officer ahead of Trump's first inauguration.

Now

President, Global Affairs — Meta's worldwide AI policy chief

Has not left Meta. Appointed President of Global Affairs January 2025. Has negotiated privacy frameworks in 50+ countries. Drives Meta's strategy on AI regulation globally — what gets regulated, what doesn't, and how.

Bush White House → Meta for 14 years → now shapes AI regulation in 50+ countries. He has never left. He is the direct line between Meta's commercial interests and the governments that regulate them.

Dina Powell McCormick

Trump NSC Meta Board

Then

Deputy National Security Advisor, Trump 1st term → Goldman Sachs VP

Assistant Secretary of State under Condoleezza Rice. Deputy NSA under Trump. 16 years at Goldman Sachs in senior leadership. No prior connection to Meta before January 2026.

Now

President & Vice Chairman, Meta — leads US Government AI infrastructure

Joined Meta's board and management team January 12, 2026. Same day, Zuckerberg announced Meta Compute — a new AI infrastructure initiative she will lead. Direct Trump White House connection at the moment AI regulation is being written.

Trump's National Security Advisor → Goldman Sachs → Meta board, January 2026. Joined 11 weeks before Zuckerberg appeared on PCAST. She is the mechanism connecting Meta's AI infrastructure to government decision-making.

Jennifer Newstead

Trump State Dept Still at Meta

Then

Legal Adviser, US Department of State (Trump 1st term)

Harvard College, Yale Law School. Served as Legal Adviser at the US State Department under Trump's first term. Joined Meta as General Counsel in June 2019.

Now

Chief Legal Officer, Meta

Oversees all of Meta's global legal affairs, AI compliance, content policy, and government engagement. The legal architect of how Meta navigates AI regulation worldwide.

Trump State Dept legal adviser → Meta's chief legal officer defining global AI compliance. The lawyer who advised the US on international law now advises the company being regulated.

Mustafa Suleyman

DeepMind Co-founder Microsoft AI

Then

Co-founded DeepMind (2010) — Google acquired for £400M

Co-founded DeepMind with Demis Hassabis and Shane Legg. Left in 2019. Founded Inflection AI ($1.3B raised). Inflection's team was effectively acquired by Microsoft in 2024.

Now

CEO, Microsoft AI

Runs Microsoft's entire consumer and enterprise AI division — Copilot, Bing AI. Reports to Satya Nadella. Microsoft is OpenAI's primary funder and cloud provider.

DeepMind co-founder → now runs Microsoft AI, which funds and hosts OpenAI. He went from building Google's AI crown jewel to running the infrastructure that powers Google's main AI competitor.

Geoffrey Hinton

Google Brain Independent

Then

VP & Engineering Fellow, Google Brain (2013–2023)

One of three "Godfathers of AI." Built the neural network foundations of modern deep learning at Google for 10 years. His students include Ilya Sutskever.

Now

Independent safety advocate — Nobel Prize in Physics 2024

Left Google specifically to speak freely about existential AI risk. "I left so I could talk about the dangers of AI without considering how this affects Google."

The scientist whose research made all of this possible left specifically so he could warn us about it. He won the Nobel Prize. His warnings are not being heeded by the people who inherited his work.

Niki Parmar &
Ashish Vaswani

Google Brain Adept → Amazon

Then

Research Scientists, Google Brain — co-authors of the Transformer

Co-authored "Attention Is All You Need" (2017) — the Transformer paper that is the foundation of ALL modern AI including ChatGPT, Claude, Gemini.

Now

Co-founded Adept AI → team joined Amazon AGI (2024)

Left Google to co-found Adept AI. Adept raised $350M. In 2024, Adept's founders and core technology were absorbed into Amazon's AGI division.

The researchers who literally wrote the paper that made modern AI possible left Google. Their Transformer architecture now powers every AI company competing against Google.

Timnit Gebru

Fired by Google DAIR

Then

Co-lead, Google Ethical AI team

Hired specifically to study AI harms. In December 2020, co-authored a paper warning about large language models. Google suppressed the paper. When Gebru pushed back, Google cut off her corporate email.

Now

Founded DAIR Institute — $3.7M in philanthropic funding

Founded the only major AI ethics research body fully independent of Big Tech. Funded by Ford Foundation, MacArthur Foundation, Open Society Foundation, and Rockefeller Foundation.

Hired to warn Google about AI risks. Fired for doing exactly that — two years before ChatGPT. She built the only independent safety institution. It has $3.7M. SoftBank's single OpenAI round was $40B.

Sundar Pichai

Google 2004– PCAST

Then

Joined Google 2004 — led Chrome, Drive, Android, then CEO 2015

Built Chrome (2008). Oversaw Android's global expansion — 3B devices. Under his leadership, Google's behavioral advertising became the model all social media copied.

Now

CEO, Alphabet — all AI strategy + Trump PCAST member

Leads Alphabet's AI-first strategy: Google DeepMind, Gemini, Google Cloud AI. Named to Trump's PCAST advisory council March 25, 2026 — same day as the addiction verdict.

Built the behavioral ad model all social media copied. Named to the AI advisory council on the same day a jury found that model caused harm to children with malice.

Demis Hassabis

Google DeepMind Nobel Prize 2024

Then

Co-founded DeepMind 2010 — Google acquired for £400M

Chess prodigy and cognitive neuroscientist. AlphaGo (beat world Go champion, 2016). AlphaFold (Nobel Prize in Chemistry, 2024). Has publicly warned that AGI risk is real and imminent.

Now

CEO, Google DeepMind — leads Gemini + all AI research

Runs Google's entire AI research and product operation. Predicts AGI within years. Has won the Nobel Prize for AI safety research while simultaneously building the most capable AI systems.

Won the Nobel Prize for research that predicts biological structures. Publicly warns about AGI risk. And is simultaneously building the most capable AI systems inside Google. This is the tension at the center of the whole story.

Neal Mohan

Google 2008–

Then

SVP Display & Video Ads (2008–15) → CPO, YouTube

Built Google's display advertising targeting infrastructure. Moved to YouTube as CPO, where he oversaw the recommendation algorithm and autoplay features central to the addiction verdict.

Now

CEO, YouTube — 2.7B monthly active users

Runs the platform a Los Angeles jury just found liable for designing addictive systems that knowingly harmed children. YouTube's share of damages: 30% ($1.8M).

Built Google's ad targeting infrastructure, then ran the platform just found liable for harming children with malice. Now oversees YouTube as AI makes its recommendation algorithm more powerful than ever.

Sergey Brin

Google Co-founder PCAST

Then

Co-founded Google 1998 — stepped back 2019, returned 2023

Co-founded Google with Larry Page at Stanford in 1998. Stepped back in 2019. Returned in 2023 specifically to accelerate Gemini development as ChatGPT became an emergency.

Now

Active AI developer at Google + Trump PCAST member

Named to Trump's PCAST advisory council March 25, 2026 — same day as the addiction verdict. Now formally advises the US government on AI policy.

Google's co-founder came back specifically to build more powerful AI — and was named to the AI policy council on the same day a jury found his company acted with malice toward children.

Prabhakar Raghavan

Google 2012–

Then

Led Search, Ads, Commerce, Maps, Shopping (2018–2024)

Ran Google's Knowledge & Information division — the entire behavioral ad and search ecosystem. Oversaw the rollout of Search AI Overviews.

Now

Chief Technologist — AI strategy adviser directly to Pichai

Stepped up from running the products to shaping Google's long-term AI strategy. Works directly with Pichai to translate AI research into strategic direction.

Ran Google's behavioral ad products for six years → now advises the CEO on AI strategy. The institutional knowledge of engagement-maximizing systems is now being applied to Gemini.

Kent Walker

Google 2006–

Then

Joined Google as General Counsel in 2006

Has overseen Google's legal strategy, government affairs, content policy, and global lobbying for nearly 20 years. Testified to Congress on AI policy.

Now

President of Global Affairs & CLO, Google & Alphabet

Shapes how AI gets regulated worldwide — what Google advocates for, what it accepts, what it fights. Advises Google's board on all legal and policy issues.

Google's Joel Kaplan — 20 years shaping the regulatory environment that Google operates in. As AI regulation is being written, he decides what Google accepts and what it fights.

Thomas Kurian

Oracle 22 yrs Google Cloud

Then

President of Product Development, Oracle (22 years)

Ran Oracle's entire product organization for 22 years. Joined Google Cloud in 2019 to build its enterprise business. Google Cloud revenue grew 35% in Q3 2024.

Now

CEO, Google Cloud — enterprise AI infrastructure

Sells AI infrastructure to governments, militaries, and enterprises worldwide. Google Cloud is directly competing with Microsoft Azure for government AI contracts.

Runs the infrastructure layer that governments use to deploy AI. As nations decide which AI systems to adopt at scale, Google Cloud — and Kurian — are in the room where those decisions are made.

The only people with no financial stake in this outcome are the only ones asking whether it is safe.

DAIR Institute — Independent AI Ethics

$3.7M

Total founding capital. From Ford Foundation, MacArthur Foundation, Open Society Foundation, and Rockefeller Foundation. Zero commercial AI investors.

vs.

SoftBank → OpenAI — Single Funding Round

$40B

SoftBank's Series F investment in OpenAI, March 2025. The largest private funding round in the history of technology.

That ratio — $3.7 million to $40 billion — is not a market inefficiency. It is a choice. The same choice we made with social media, made again, faster, with technology that is more capable, more intimate, and more consequential than anything we have built before.

I joined The Facebook on January 2, 2005. I left every Meta platform in September 2025 after reading Careless People. I had been on the platform for twenty years. This is a moment that demands collective pause. Because the careless people didn't go away. They went to OpenAI.

The People Who Built the Addiction Machine Now Run AI

The People Who Built the
Addiction Machine
Now Run AI