Public transcript

OpenAI's Sam Altman trusts AI will treat most diseases by 2035

By: CNN
Length: 24 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 CNN interview, Sam Altman discusses AI's impact on jobs and argues government should be doing more research in AI, longevity and fusion. He explains his investment in Retro, a company pursuing partial cellular reprogramming to add 10 years of health span, and predicts AI will treat or mitigate most diseases by 2035. He also shares concerns about AI and mental health, a personal GLP-1 overdose story, and his skepticism about brain uploading and longevity fads.

Key points

Sam Altman says some jobs will go away with AI but new ones are created quickly after every technological revolution.
Altman invested in Retro, a company pursuing partial cellular reprogramming with a goal of adding 10 years of health span, not immortality.
He predicts AI will be able to treat or mitigate most diseases by 2035.
GPT-5 made a big push to improve healthcare queries, and health is one of the biggest uses of ChatGPT.
Altman is increasingly worried about AI's effects on societal-scale mental health as people use it as a therapist and life coach.
He believes people will significantly prefer a human doctor even if an AI doctor performs better.
Doctors secretly use ChatGPT constantly and ask for a HIPAA-compliant version for work.
Altman took a mega overdose of a GLP-1 from a compounding pharmacy and ended up in the ER.

Questions it answers

00:00Jobs and government's role

What does Altman think about AI's effect on jobs and the government's role?

He says some jobs go away with every technological revolution but new ones are created quickly, and argues government has abdicated its responsibility — in a well-functioning society OpenAI would have been a government effort.

New jobs follow technological revolutions quickly
Altman believes government should be doing AI, longevity, and fusion research

01:11Longevity and Retro

Why is Sam Altman interested in longevity and what is Retro?

He was drawn by the idea that most diseases are diseases of age, and by partial reprogramming — making cells somewhat younger — so he made a large investment in Retro, aiming to add 10 years of health span.

Partial reprogramming rewinds cellular age somewhat
Retro's goal is 10 years of health span, not immortality

04:48AI in healthcare

How is AI being used in healthcare today?

Healthcare is one of the biggest uses of ChatGPT, GPT-5 made a big push on health queries, and he imagines an AI medical advisor that flags risks from sleep, diet, blood tests and symptoms.

GPT-5 improved health-related queries
Doctors secretly use ChatGPT constantly and want a HIPAA-compliant version

07:02Mental health concerns

What worries Altman about AI and health?

He is increasingly worried about societal-scale mental health as people use AI as a therapist and close advisor, and says OpenAI is working hard to understand it.

ChatGPT is only about three years old and usage shifted fast
Societal-scale behavior could change societal-scale mental health

09:582035 prediction and superintelligence

What does Altman predict AI will do for disease by 2035?

He predicts AI will treat or mitigate most diseases by 2035, and believes a vastly smarter intelligence could figure out complex biology and explain it to humans.

AI both speeds discovery and makes new insights
He defers mostly to biologists and pharma companies

14:05GLP-1 overdose story

What happened with Altman's GLP-1 overdose?

A compounding pharmacy messed up his GLP-1 dosing, causing a mega overdose, vomiting, a pulse around 150, and an ER visit; he also criticizes charlatans selling longevity fads.

He had no tolerance for the drug when overdosed
He finds ice bath and fasting enthusiasts smug and annoying

19:02AGI, ethics and prompts

How does Altman view AGI, ethics and the right prompt?

He sees AI as a tool, not a self-aware creature, hopes AI helps humans devise better win-win ethical systems, and says the prompt people want is 'help me flourish in accordance with my goals.'

AGI is a revealing Rorschach test for people
Ethics already runs behind, but AI may help generate better systems

22:02Uploading, death and personal views

Would Altman upload his brain or merge with AI, and how does he want to die?

He is very skeptical of brain uploading and merging with AI, likes his body and the world, wants to die quickly and painlessly when very old, and hopes to stay healthy well past 80.

He sees horrible ways brain uploading could go wrong
He assumes he'll live much longer than 80 if things go well

Notes

Jobs and Government

Altman says some jobs always go away with technological revolutions, but new ones are created quickly afterward.
He argues government has abdicated its responsibility in science research; in a well-functioning society, OpenAI would have been a government effort, and government should be doing AI research, longevity, and fusion.

Longevity and Retro

His interest stems from the observation that most diseases are diseases of age; 20-year-olds don't get nearly as sick as 80-year-olds.
Partial reprogramming means rewinding cellular age somewhat — not back to a stem cell — making cells a bit younger.
He made a large investment in Retro, whose goal is adding 10 years of health span; it may not extend life at all, just keep you healthier longer.
He has become increasingly negative on academia's ability to make things happen beyond early discovery.

AI in Healthcare

Healthcare is one of the biggest uses of ChatGPT; GPT-5 made a big push to improve health-related queries.
He predicts AI will treat or mitigate most diseases by 2035, citing the Retro research project as pushing research years ahead.
Doctors he visits say "don't tell anyone, but every doctor here uses this all the time" and asks for a HIPAA-compliant version.
He distinguishes consumer health information from using AI to accelerate longevity research — different approaches, same intelligence.

Concerns

He is increasingly worried about societal-scale mental health as people use AI as a therapist and close advisor.
He is sure people will significantly prefer a human doctor even if the AI performs better; he personally would want a human doctor for serious issues.
He sees charlatans making money in longevity because people want a magic pill instead of sleep, exercise, and moderation.

Personal Stories

A compounding pharmacy screwed up his GLP-1 dosing, causing a mega overdose, a night of vomiting, a pulse around 150, and an ER visit.
He finds ice bath and fasting enthusiasts annoying and smug.
He is skeptical of brain uploading and merging with AI; he likes his body and the world.
He wants to die quickly and painlessly, like a stroke or heart attack in his sleep when very old.

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