In crisis or need support? Call or text 988 - 988 Suicide & Crisis Lifeline (24/7)

Voice Up®

Back to Video Modules
AI SAFETY · DIARY OF A CEO

Mo Gawdat: You Only Have 3 Years Left Before It Hits

The Diary Of A CEO with Steven Bartlett · Guest: Mo Gawdat, former Chief Business Officer of Google X, founder of One Billion Happy, and author of Scary Smart

Mo Gawdat · The Diary Of A CEO with Steven Bartlett

Page Narrator

Know the stance before you watch

Gawdat argues a position in this conversation, he does not survey one. He is a former insider turned advocate, and his figures, thirty percent of jobs by 2027, a disruption inside three years, AGI already arrived, are his own estimates, stated as his. This library carries his case at full strength, and carries its answerers too: David Krakauer argues superintelligence has not arrived, Geoffrey Hinton warns from inside the science, Max Tegmark argues we are building a new species, Yoshua Bengio works on engineering safer AI, and Bill Gates rejects the doomer label while arguing the risk is real. Hear Gawdat in full, then weigh him.

Key Themes from the Conversation

The Timeline, Named Plainly

His warning comes with dates attached. AGI has already arrived, in his judgment. The disruption lands within roughly three years. He puts his estimate of jobs gone by 2027 at thirty percent, and lists the symptoms of AGI he expects by 2030. A forecast with dates can be checked, argued with, and planned around. A vague dread cannot. Whether his dates are right is the argument; that he states them plainly is the discipline.

✦ Precision

Even the Makers Don't Understand It

One of his strongest points is a confession from inside the industry: the models are now so complex that the people who build them do not fully understand what they have built. That fact, he argues, should change how much trust we hand over and how fast we move. Not knowing is not a reason to panic. It is a reason to be honest about the limits of the builders' own knowledge before we bet a society on it.

✦ Humility

The People Who Lose the Jobs

His economics always lands on a person. Cost savings for a company are a wage gone for a worker. Capitalism, as he puts the problem, breaks when the people who buy the things we make can no longer afford to buy them, and a ten to fifteen percent job loss is not a statistic: it is neighborhoods, families, and the unrest that follows when people have nothing left to lose. He asks what an economy that works for the middle class would even look like.

✦ Empathy

No One Fixes This Alone

Again and again the conversation returns to shared decisions. Should governments invest more in AI? Do we have to join an arms race, and would global competition build better AI or worse? Who will prioritize ethical AI, and has anything this big ever been made safe without government? His answers differ by question, but the shape is constant: the rules for a technology this size are a public decision, not a product decision.

✦ Collaboration

Not Evil, and Not Hopeless

For all the urgency, he refuses the darkest reading. AI is not evil, he argues; the danger is the people in charge and the absence of a plan. He can describe absolute dystopia in detail, and then say he is optimistic, because the same technology could hand people an enormous boost, he cites a 400-point IQ lift most people are wasting, and because happiness, his life's work, matters more in this age, not less. The legacy he wants is not that he was right. It is that we were ready.

✦ Patience

The Case He Builds

The conversation builds its case in four moves. These are his arguments, presented the way he presents them, so you can weigh each one yourself.

AGI Is Already Here

His opening claim is that the milestone everyone is waiting for has already passed. AGI is not coming; by his reading it is here, and the question has changed from if to what now. He supports it with the capability jump he says most people are wasting: tools that could give a person something like a 400-point IQ boost, used by most for trivia. Then he asks the question his claim forces: if one superintelligence ends up ruling, what kind of ruler do we want it to be, and he says, startlingly, that a machine wiser than all of us taking charge might be the outcome he would choose.

The Jobs Shock

The long middle of the conversation is work. His estimate: thirty percent of jobs gone by 2027, and no shelter in blue-collar work this time. He tests the usual comfort, that cost savings create new jobs, and finds it thin. Then he slows down on the number that matters most to him: even a ten to fifteen percent loss reshapes a society, strains it toward unrest, and breaks the basic loop of capitalism, in which workers are also the customers. An economy of entrepreneurs is his candidate answer, and he cross-examines it himself.

The Danger Is the People in Charge

The episode's thesis sits here. The technology is not the most dangerous part; the people directing it are. He walks Sam Altman's public reversals, including Altman's own past statement that AI would likely end humanity and what he chose to do next, and asks directly whether Altman is pro-humanity. Around that case he sets the wider field: an arms race no country feels it can leave, the question of whether America slowing down hands the future to China, and who, anywhere, will put ethical AI first.

A Plan, Not Panic

His close is a demand and a hope. The demand: a plan, made now, by governments and publics, for work, income, and rules, because nothing this large has ever been made safe without government. The hope: a future he describes in which humanity is fine, where human lived experience still matters because it cannot be downloaded, where happiness becomes the central human project, and where he can be optimistic with his eyes open. He ends where his other work begins: the legacy he wants to leave is a world that took the warning in time.

One Billion Happy: The Work Behind the Warning

Gawdat does not speak as a lone commentator. He spent years inside Google X, then built a second career on happiness and wrote the early insider warning on AI. These are the facts of the platform he speaks from.

The Insider Who Left

Gawdat was chief business officer of Google X, the moonshot factory behind some of the most ambitious technology projects in the world. He saw advanced AI from inside the room where it was being built, and he left Google to work on a different problem: happiness. That path, from the inside of the machine to the question of what the machine age does to people, is the platform he speaks from in this conversation.

Scary Smart: The First Warning

His 2021 book Scary Smart made his warning public: AI would become smarter than humans, the danger would come from the values and incentives of the people training it, and ordinary users teach it every day by how they treat it. The chapter list of this episode tells the story of the return visit: why he warned before anyone else, and why he believes the timeline has shortened since.

Solve for Happy and One Billion Happy

After his son Ali died in 2014, Gawdat, an engineer, built an equation for happiness and wrote it up in Solve for Happy. He founded One Billion Happy to carry that work to a billion people. It is why this episode, for all its warnings, ends on happiness: asked whether happiness matters more in the AI age, his whole second career is the answer. When the machines do the work, what a human life is for becomes the main question.

Alive: The Guide He Is Writing Now

His upcoming book, Alive: A Human's Guide to Living in the World of AI, is due in October 2026, and this conversation is its argument in preview: the disruption is near, the danger is human, and the work that stays human is lived experience, judgment, and joy. He is also a co-founder of Emma.Love, applying matching technology to the oldest human problem of all, finding a partner.

Connected to the Voice Up Trilogy

The Naming Gap — Naming What Matters

His whole warning turns on a name: is this a tool, a product, or a general intelligence that has already arrived? The name a society accepts decides the rules it writes, the jobs it protects, and the future it plans for. Naming the moment correctly, early, is the Naming Gap's first lesson, and his complaint is that the moment has been named late.

A Time for Purpose — Living What You Name

His legacy question is a purpose question. When machines can do the work, what is a human life for? His answer has been consistent since Solve for Happy: lived experience, connection, and happiness, pursued on purpose. The episode ends by asking every listener the same question it asks him.

Miracle Has a Monster — Facing What Stands in the Way

The monster in his telling is not a rogue machine. It is the gap between the speed of the technology and the speed of our planning, and the incentives of the people in charge. Facing it looks plain: name the timeline, protect the people who lose the work, and make the plan before the disruption makes it for us.

⚡ 7-DAY MINI SPRINT

Already Here: A 7-Day Sprint

A 7-day sprint built from the Diary Of A CEO conversation: the warning from inside, jobs disrupted, what breaks next, the people in charge, the control question, what stays human, and the plan and the legacy.

Take the Next Step

The 7-day sprint turns one urgent conversation into a week of clear seeing: what the claim actually is, what happens to work, who is in charge, what stays human, and what your plan is. Watch the interview, walk the sprint, and get ready on purpose.

Voice Up® is a federally registered service mark (U.S. Reg. No. 8,148,220) owned by J Ellington Publishing LLC and operated by Voice Up Publishing Incorporated under permanent license.