Public transcript

Sam Altman's BIG AI PREDICTION: the greatest BOOM in GLOBAL COMMERCE | FULL G20 SPEECH

By: Diario AS
Length: 32 min
Language: English
Text from: the video's captions
Transcribed: 13 October 2026

Read the full transcript Open on YouTube

Notes in English.

Summary

In this G20 lunch conversation, Sam Altman traces his path from AI enthusiast to OpenAI's founding at the end of 2015, through the deep learning breakthrough of 2012, GPT-1 in 2018, GPT-3 in 2020, and the GPT-4 moment in early 2023 when the technology's potential became clear. He argues AI will drive the greatest boom in entrepreneurship and small business creation the world has ever seen, and that adopting AI is

Key points

Sam Altman describes himself as a lifelong AI and sci-fi nerd who gave up on AI in school, then returned to it after the 2012 deep learning breakthrough and co-founded OpenAI at the end of 2015.
Altman says the moment it became clear AI was "bottling knowledge" was around GPT-4, which OpenAI had internally for about eight months before its March 2023 launch.
He predicts AI will be "the greatest boom in entrepreneurship and the creation and development of small businesses the world has ever seen," citing a laundromat owner who ran his business with ChatGPT.
Altman calls AI adoption "non-negotiable" for countries, comparing rejecting AI to rejecting electricity a century ago.
He frames the history of technology as one continuous exponential built by successive generations, not isolated revolutions, and warns that countries that miss out will not be ready for what comes next.
He differentiates OpenAI as "pragmatic centrists" who want to be a reliable platform, avoid catastrophic risks, respect national sovereignty, and stay "on team humanity."
He cites massive token growth—from the top user at 100,000 tokens a month in early 2020 to hundreds of billions by mid-2026—to argue that vast compute infrastructure must be built to keep AI abundant and inexpensive.
Key risks he names are cybersecurity, biosecurity, and too much concentration of power, and he argues worry itself is how society anticipates and mitigates technology's downsides.

Questions it answers

00:08Origin story

How did Sam Altman go from being a sci-fi nerd to leading OpenAI?

Altman loved AI as a kid but gave up on it in school when nothing worked. After the 2012 deep learning breakthrough, he concluded AI was going to happen and started OpenAI at the end of 2015, with work beginning in January 2016.

The 2012 deep learning watershed showed systems improved with more compute.
OpenAI's first day of work was January 2016.

02:34The GPT-4 moment

When did it become clear that AI was bottling knowledge?

Altman points to the GPT-4 period, around March 2023, after OpenAI had the model internally for about eight months. Even internal skeptics concluded the approach would go very far.

GPT-3 in 2020 was the first model over a usability threshold.
Scaling laws suggested applying massive compute would produce something amazing.

05:57Entrepreneurship boom

Why does Altman think AI will transform small business creation?

He predicts the greatest boom in entrepreneurship ever, because people with ideas but no capital or experts can now use AI for marketing, contracts, supply chains, and science. He cites a laundromat owner who ran his business with ChatGPT in 2023.

Coding and white-collar agents accelerated this in 2025.
Long-term persistent agents acting as virtual collaborators are coming next.

08:54Platform, not context

Will OpenAI provide the context for how AI is used?

No. Altman says OpenAI provides the engine or commodity of AI, while the world's people must supply the context, questions, and ideas. He says OpenAI wouldn't be good at that role even if it wanted it.

OpenAI wants to respect national sovereignty in how countries use AI.
Altman compares it to electricity: he may not like every use, but everyone should have it.

11:31One continuous exponential

How should governments think about AI compared to past technologies?

Altman rejects treating revolutions as isolated events, arguing history is one continuous exponential where each generation builds scaffolding for the next. Countries that don't enable their people with AI won't invent the next layer.

There will be bigger revolutions after AI.
Missing any part of the continuum leaves a country unprepared.

14:22What differentiates OpenAI

What is OpenAI's DNA compared to other frontier companies?

Altman calls OpenAI the pragmatic centrists, avoiding blind optimism and doomerism, wanting to be a reliable platform that avoids catastrophic risks while respecting sovereignty. He stresses the company is proudly on team humanity and doesn't want to be the only player.

OpenAI doesn't want to suck up all the value.
People should stay at the center of the economy and run the future.

19:06Non-negotiable adoption

What does Altman mean by AI adoption being non-negotiable?

He compares rejecting AI to rejecting electricity a century ago. Countries can regulate differently and choose to build or rent data centers, but every person, business, and society will need AI to deliver the quality of life citizens demand.

In a decade, people will expect AI everywhere and stop talking about it.
Kids growing up today will never know a world without smart products.

21:58Risks and abundance

What could hold the world back in five years?

Altman names cybersecurity as an urgent challenge where things will go very wrong without action, plus biosecurity and the risk of concentrated power. He argues AI must be made abundant and inexpensive, or it becomes something rich people compete over, worsening inequality.

Token use has grown multi-million-fold in six and a half years.
He compares AI abundance to the collapsing cost of an hour of light at night.

Notes

Sam Altman's Path to OpenAI

Altman describes himself as a lifelong AI and sci-fi nerd who studied AI in school, gave up when nothing worked, and pursued entrepreneurship instead.
The 2012 deep learning breakthrough changed everything: systems improved with more compute, and by 2015 he concluded AI was going to happen.
OpenAI was announced at the very end of 2015, with the first day of work in January 2016.

Key Milestones

Early work included training a robot hand to solve a Rubik's Cube and playing video games.
Progress on unsupervised learning came around 2018, producing GPT-1; GPT-3 followed in 2020 as the first model over a threshold of usability.
The consensus moment that AI was "bottling knowledge" came around GPT-4, which OpenAI held internally for about eight months before launching it in March 2023.

Predictions for the Global Economy

Altman predicts the greatest boom in entrepreneurship and small business creation ever, illustrated by a laundromat owner who ran his entire business with ChatGPT in 2023.
2025 brought acceleration via coding agents and white-collar work agents; he expects long-term persistent agents that act like virtual collaborators next.
He calls AI adoption "non-negotiable" for countries, comparing it to electricity: some may build data centers, others may rent them, but all will need the commodity.

OpenAI's Position

He calls OpenAI the "pragmatic centrists," avoiding both blind optimism and doomerism, and says the company is "on team humanity."
OpenAI sees itself as providing the engine; the world's people must supply the context, questions, and ideas.
He says OpenAI does not want to be the only company or absorb all value, as that would harm global stability.

Scale and Infrastructure

The top token user went from roughly 100,000 tokens a month in early 2020 to hundreds of billions by mid-2026; global per capita use went from zero to 100,000 tokens a month.
He projects average use could reach 100 billion tokens a month within another six and a half years, requiring massive infrastructure investment to keep AI abundant and inexpensive.
He compares this to the falling cost of an hour of light at night, which once took five hours of labor.

Risks and Outlook

Top risks: cybersecurity (where he says some things will go very wrong without urgent action), biosecurity, and concentration of power.
He argues worry is how society anticipates problems, citing how fire safety was developed after cities burned down.
He remains optimistic, noting AI safety has gone better than doomers predicted five years ago, and shares a story of someone whose life was saved by a ChatGPT diagnosis.

Text from the video's own captions. Summary and notes written by AI from the transcript, so check anything important against the recording.