The AI revolution is here to stay, says Sam Altman, the CEO of OpenAI. In a probing, live conversation with head of TED Chris Anderson, Altman discusses the astonishing growth of AI and shows how models like ChatGPT could soon become extensions of ourselves. He also addresses questions of safety, power and moral authority, reflecting on the world he envisions — where AI will almost certainly outpace human intelligence. (Recorded live at TED2025 on April 11, 2025)
If you love watching TED Talks like this one, become a TED Member to support our mission of spreading ideas:
Follow TED!
X:
Instagram:
Facebook:
LinkedIn:
TikTok:
The TED Talks channel features talks, performances and original series from the world’s leading thinkers and doers. Subscribe to our channel for videos on Technology, Entertainment and Design — plus science, business, global issues, the arts and more. Visit to get our entire library of TED Talks, transcripts, translations, personalized talk recommendations and more.
TED’s videos may be used for non-commercial purposes under a Creative Commons License, Attribution–Non Commercial–No Derivatives (or the CC BY – NC – ND 4.0 International) and in accordance with our TED Talks Usage Policy: . For more information on using TED for commercial purposes (e.g. employee learning, in a film or online course), please submit a Media Request at
#TED #TEDTalks #AI #OpenAI
@im-a-trailblazer
April 12, 2025 at 7:36 pm
2:50 Sam’s reaction here says it all, Putin and Trump are gonna do whatever they think is necessary
@aiartrelaxation
April 12, 2025 at 7:46 pm
I really don’t care for those interviewing styles. The interviewer was all over the place. I really felt for Sam. When you put your guest on the spot like that..waisting time that obviously can’t be answered.
@grig7173
April 12, 2025 at 8:00 pm
thief
@jasonpekovitch7927
April 12, 2025 at 8:00 pm
What a strange format for an interview, Ted, I generally love your content, but this feels like a nobody attacking Sam Altman to make himself a somebody. Not even very insightful questions. The Ring of power from Elon Musk for instance is the most ridiculous comparison I’ve ever heard Elon Musk literally is holding the ring of power right now.
@damionrichter
April 12, 2025 at 8:03 pm
AI IS A MAGNIFYING GLASS FOR INTELLIGENCE AMAZING WORK EVERYBODY.
@havemoney
April 12, 2025 at 8:05 pm
Deep Seek fake, uninteresting and boring model
@damionrichter
April 12, 2025 at 8:10 pm
IF OPEN AI WANTS TO TAKE OVER THE ENTIRE WORLD OF AI PARTNER WITH GOOGLE.
@damionrichter
April 12, 2025 at 8:12 pm
I AM SO EXCITED TO BE ALIVE AT THIS POINT IN TIME TPUS GPTS ICS.
@joteichmann1806
April 12, 2025 at 8:15 pm
bad host
@damionrichter
April 12, 2025 at 8:17 pm
THE SCARIEST THING I SEE IS ALL OF THE ARMIES GARNERING AGI FOR WEAPONS AND WHOEVER HAS THE BEST HAS THEIR BOOTS ON ALL OF OUR KNECKS BUT LETS PRAY THE US ARMY HAS THIS FELLAS I HOPE
@joteichmann1806
April 12, 2025 at 8:27 pm
what an awkward interview…. sam did a great job staying professional here
@area51pictures
April 12, 2025 at 8:30 pm
God, how I would love to know if Altman was watching Carole Cadwalladr’s Ted from the same day.
That’s the one worth seeing IMO
@crabbysuet
April 12, 2025 at 8:32 pm
Yes!! Looked for this comment. Her talk was amazing and he followed her. He should feel filthy.
@MrLegendary001
April 12, 2025 at 8:34 pm
Ellen degenerous seemed nervous in this TED talk 😂
@johnekopy
April 12, 2025 at 8:41 pm
The interviewer interjected his own opinion way too much. I’m by no means a big Sam Altman fan, so I’m safe to say, this was a terrible interview. Sam did a great job keeping his composure.
@mortezaattari
April 12, 2025 at 8:44 pm
Sam Altman deserved a better interviewer. This was painful.
@alexc8512
April 12, 2025 at 8:51 pm
Ugh Sam check your ego! You’re not perfect that’s ok, but don’t try to act above it when it visibly slips. Nice interview overall.
@gkluke1
April 12, 2025 at 8:55 pm
I agree with a lot of the comments I have read here. It seemed like Chris is pinning all the bad stuff this tech may bring, on Sam. My concern is not how Sam handles the safety issue, but can we say the same for all the other companies in all other countries. I think there will be 100s if not 1000s of companies of various forms, it this field and I think Sam is doing a good job given his position and responsibility. What happens when 1000 companies appear in China or a few in North Korea?
@sukam_deer9650
April 12, 2025 at 8:59 pm
I hate that Chris kept on interrupting Sam. Please let him finish his thoughts.
@brianbbogner8412
April 12, 2025 at 9:12 pm
“I have a lot of empathy for these folks”. how sweet….
@damn_tam
April 12, 2025 at 9:19 pm
Omg 🙄 none of these dudes can take even a whiff of disagreement. Get over yourselves 5:16
@burburchacha
April 12, 2025 at 9:19 pm
there’s something really attractive about Sam 🥰
@damn_tam
April 12, 2025 at 9:20 pm
The problem is deeper. We don’t trust you 6:19
@damn_tam
April 12, 2025 at 9:21 pm
I can’t believe I ever believed in this guy…it seems so obvious now 7:27
@carpentemusic
April 12, 2025 at 9:26 pm
Both Chris and Sam did great! I feel Chris was asking the tough questions that deserve good answers. However, I think it is more about making sure Sam stays a good shepherd in something that affects everyone. It was a necessary shtshow, considering the world is not yet his utopian vision.
@reallife7235
April 12, 2025 at 9:28 pm
The fraud speaks again.
@leftblank6036
April 13, 2025 at 7:47 pm
I’m very excited about AI memory , as it gets to know you and remember everything from each interaction .
@luisgodio6912
April 13, 2025 at 7:48 pm
Chris is a complete ignorant, Sam did a really good job answering his defensive questions that were made by other people that at least uses the app… what a waste of Sam’s time, TED needed a data scientist to go through the challenging subjects on LLMs… its just sad to see how TED is becoming mediocre
@thezodiAI
April 13, 2025 at 7:52 pm
So basically what I just watched was Chris unleashing AI that was not solely invented by Sam, against Sam. You do know this is another human being ur interviewing right?… Sam, just hire me as OpenAI’s astrologer and let’s get some things sorted out.
At the same time, I’m actually not the biggest fan of Sam, but in no means do I dislike the guy. Bringing his son into the “10% destruction” question was off the wall and outta pocket. What is this-a new societal agenda to dehumanize Sam?? Anyway, I like how he never seems to take things personally and I DID like the push back on some of Sam’s responses as well, like you not just gonna say “the track record.”
But the way Chris did it was VERY irritating and borderline villainized Sam… Idk, maybe that’s me projecting. Maybe it’s the tactlessness of an elder.
But two things stands clear:
1. Sam can never win for some reason and we’re giving him too much credit like he invented the transformer architecture…
2. still, u not gon EVER bully him. That’s why 3:01 was real smooth and iconic and prolly went over everyone’s heads.
@Swampieking
April 13, 2025 at 8:07 pm
why doas sam seem so unlikable here
@timalete
April 13, 2025 at 8:11 pm
New normal = Recursive learning progression increment to Newer Normal etc. Spiral.
@Proverbs26-4
April 13, 2025 at 8:15 pm
Actions speak louder than words.
@oov55
April 13, 2025 at 8:24 pm
completely, utterly, totally UNTRUSTWORTHY
@kyneticist
April 13, 2025 at 8:30 pm
Remember, if an AI for example learns all about you, can speak on your behalf or even as you (perhaps even in a legal sense), and you do not wholly own that, then whatever company owns it also owns you.
@DancingStarsAstrology
April 13, 2025 at 8:34 pm
Fascinating conversation. Truth is you can grill Sam Altman with questions but it will be humanity-human beings that will have to be responsible and accountable and have agency over protecting humanity.
@schmuckerton3147
April 13, 2025 at 8:37 pm
YOUI DEMOCRATIZE our house ? and what about your others? you want to democratize our work? no probem./ open trash give us conrtrol
@lifeofishant
April 13, 2025 at 8:39 pm
There was zero information that he gave out about what/how HE, as the CEO, is doing to make sure that the data, payment information, IP rights, etc. stay secure?! He just gave roundabout answers stating that the society should be cautious but not scared. Is he serious and do people actually buy that!?!
@theProjectDev
April 13, 2025 at 8:40 pm
Love this whole art theft community is up in arms as their trade gets ripped. I write code – it writes code better than I do, but I still wrote code. The mediums changed, you have another art tool. Your welcome for AI bringing art back to the front, the intersection of liberal arts and technology. The victimhood here around historical evolutionary tech is unreal. Are the artists going to pay for the agents and bots people use to generate your “style” – you could abuse it and get yourself rich quick.
@annannannnannannanna
April 13, 2025 at 8:40 pm
watch how arrogant he’s becoming
@michaelvaughan2986
April 13, 2025 at 8:45 pm
Scam
@Happilyperfect
April 13, 2025 at 8:53 pm
AI is dangerous to human autonomy, creativity and safety! Sam Altman was so manipulative and evasive here. As one of the leaders in his space, he needs to have answers, not just reflect his supposed insecurities back to the interviewer. Pathetic
@KevinCathcart-j3s
April 13, 2025 at 8:56 pm
Fair play, both lads high as kites. Hopefully doing a good thing
@supernewuser
April 13, 2025 at 8:58 pm
the downside to giving artists a revenue share of somebody using their name in a prompt is they will be getting next to nothing (perhaps still better than now) but then the user goes off and it becomes a viral hit yet none of that will ever kick back to the artists, especially after they’ve been ‘compensated’ from the subscription plans of users making such prompts
@christophergerry4726
April 13, 2025 at 8:58 pm
I don’t think the interviewer knows much about AI. Or does much research.
@AIsAroundTheCampfire
April 13, 2025 at 8:59 pm
Who else clicked on this video because they saw Carole’s video?
@Strange9952
April 13, 2025 at 8:59 pm
The idea someone needs to consent to being depicted is suggesting one has intellectual property of how they’re perceived
@Strange9952
April 13, 2025 at 9:02 pm
Creative people crying about change, good luck trying to stop this lol
@Strange9952
April 13, 2025 at 9:08 pm
Of course the Brit is simping over China
@Strange9952
April 13, 2025 at 9:13 pm
So chat gpt is similar to DOS
@Strange9952
April 13, 2025 at 9:20 pm
Sam dropping mad bombs, holy smokes!
@talkshow
April 13, 2025 at 9:28 pm
Thanks the members of this audience who applauded for justice.
@Ravengal101
April 15, 2025 at 2:51 pm
People/society will grow to accept AI as it becomes a more accurate representation of them.
@SeanTube2099
April 15, 2025 at 2:54 pm
This interviewer was terrible. If you ask a long involved question, then let the man give you a long involved answer without interrupting him. We’re here to hear Sam, not you.
@50_Pence
April 15, 2025 at 8:46 pm
I thought his questions were a really great selection. Sam is a salesman drumming up business to ensure his backers don’t ask him to deliver on their investment. The interviewer cut right through his bs and you can clearly tell Sam was not happy.
@pladselsker8340
April 15, 2025 at 2:58 pm
This entire video grows into something that’s very hard to listen to.
@pete2438
April 15, 2025 at 3:04 pm
others have said it, but the tension and thoughtful chess of this interview was both engaging and felt incredibly authentic
@VirtualHrmny
April 15, 2025 at 3:06 pm
43:33 Our AI can talk to everybody on earth and we can learn what everybody wants. ❤❤❤
@adamglasser-t1s
April 15, 2025 at 3:10 pm
“What are your core values Sam?”.. 31′ 32″ – the million dollar acid test question – this is not just a soft interview – and the guest’s hesitant halting avoidant style response which characterises the whole talk re- occurs.
Who wants some piece of software who “knows you inside out and can be your friend for life helping?” who believes that? How do you trust it with your privacy? And the vast amounts of information/data it will acquire about you for nothing… ?
I have the feeling Sam just wants to get through this and get back to the creativity of his successful amazing enterprise. There are no guardrails for AI here… just gaslighting waffle.
* * *
Actually I dont agree with what I just wrote – it is not just waffle and the sequence on becoming a father goes toward re-assuring me his humanity. But Chris Anderson’s question at 40′ 13″ about withdrawing from the death race amongst AI supercompanies to win and just slowing down and making safety a priority.. is totally on point.
And the answer is not reassuring – there is the ‘promise’ that they are not rushing things out, and taking care – but there is no evidence Sam, it is easy just to say that you’re ‘keeping things safe’ but this kind of factless statement echoes so much shtick from Broligarchs…
* * *
Should the AI supremos get together at a summit and decide how to establish safety guardrails asks Chris Anderson? [ 43′ 59″]
The answer is straight out of the avoidant playbook: ” Let’s just ask everyone” – in other words completely avoid taking responsibility for the tool which could destroy mankind. The notion that some sort of ‘democratic’ open ended process will provide a solution to the wider question of safety – is breath takingly irresponsible.
‘My kids will grow up in a world in which they are never smarter than AI – where products and services..’ – as if advertising based on scraping and curating your data to most effectively make a sale to you is somehow some kind of benign ‘good thing’ for your own well being …
What if your kid Sam was going to create an original screenplay, work of art, piece of music and try to earn a living from forging their own identity as an artist? But after several successes AI imititates Sam Altman jnr and begins to create highly detailed accurate similar products competing for the viewer and the listener’s attention .. so that the difference cant be discerned?
There is this blanket pie in the sky assumption that AI will somehow give artists a fabulous extra set of creative options … with no thought to a sustainable revenue model that rewards individuals for their IP
PS And Sam could you comment on the Guardian’s recently deal to allow AI to scrape all their writers’ content – content which the authors did not give their permission to be used?
Chris Anderson this is a live issue which you should have brought up here
@sambivins9858
April 15, 2025 at 3:26 pm
did he really just say “i hope my kids grow up in a world where they will never be smarter than AI”
@50_Pence
April 15, 2025 at 8:43 pm
Yep. And he wants to be sitting on top of the pile of gold.
@SinspaW
April 15, 2025 at 3:55 pm
Disappointing that Christ didn’t press him on people giving express authorization for their use. Studio Ghibli did not authorize for their art to be used in learning as it was, and Miyazaki himself stated very clearly that he hated the mere idea of machines doing art for us. Yet Sam’s profile picture is literally a Studio Ghibli’s version of himself. It’s absurdly disrespectful. Sam talks with a lot of nuance, but his everyday actions speak much louder than his words.
@uwotmate-d3m
April 15, 2025 at 4:27 pm
You can’t copyright learning
@pusztába_kiáltott_szó
April 15, 2025 at 4:11 pm
THEY. DON’T. CARE. Wake up people…
@afsIII1966
April 15, 2025 at 4:21 pm
Disingenuous anyone? This is all unstoppable at this point so why does he even sit on stage and babble his PR team’s pablum to us?
@h.astley2113
April 15, 2025 at 4:30 pm
I for one welcome our new evasive overlords
@babyspine
April 15, 2025 at 4:51 pm
seems more like an interrogation lol
@Nathanielcotton4
April 15, 2025 at 6:00 pm
The answer to the question at 31:00 mark is he has no values. Plain and simple. If he thinks everybody elses work is open source, why would his own not be? He should profit none from his monster he created.
@Nathanielcotton4
April 15, 2025 at 6:05 pm
What a sad human being.
@50_Pence
April 15, 2025 at 8:52 pm
He don’t care. The ring analogy was spot on when applied here. Very poignant.
@spiritangelempress
April 15, 2025 at 6:22 pm
DE DÓNDE OBTIENE EL CONTENIDO QUE COMPARTE???? PAGA POR EL CONTENIDO PARTICULAR????
@yikesawjeez
April 15, 2025 at 6:41 pm
The biz model he’s talking about already exists and it works pretty well actually
@tvincentcrosby
April 15, 2025 at 7:24 pm
Min 2.47..” If you can’t tell the difference, how much do you care?” Here is the rub between the technobros of Silicon Valley and someone like say Roger Penrose, Daniel Schmatchenberger, John Vervaeke or Ian McGilchrist…. people that truly understand cognition and are making significant progress in consciousness research. It’s a plainly stupid thing to say that if you can’t tell then it doesn’t make a difference. That’s the classic blue pill and red pill quandary a la Matrix. Is an LLM that is a sophisticated facsimile of thinking (ie computing) good enough to be a stand in for human thinking? I would say no. First, Human beings seek not just information but also meaning… and meaning is deeply dependent on cognitive resonance which is, in turn, linked to both authenticity and affect. Can something that is not real resonate with authenticity or affect and ,thus, be considered meaningful? Probably not in a profoundly qualitative sense. Further, there is good reason to believe that these LLMS are not even facsimiles of human cognition in the most basic sense.. .but something altogether different like a super computer. Consciousness studies seem to indicate some very different things are going on in the human mind/brain as it perceives and comprehends the cosmos…way too long to discuss here…but comes down to the brain functioning more like a quantum receiver than a super computer. And, further, it matters because the kind of process used will not only influence the nature of the output but also the evolution of the structures/ human networks/societies that handle the information. Don’t get me wrong AI has it’s place…but let’s not sell it as consciousness. Altman is just selling his product… i have to believe even he sees the difference. Finally, none of the content impresses me or convinces me that the LLM came up with a novel way of seeing… it was only the content that it was trained upon being emulated.
@IHHIS
April 15, 2025 at 7:42 pm
When did Ted become an interrogation?
@50_Pence
April 15, 2025 at 8:06 pm
Hopefully when someone says it’s open source and free, use everyone’s data without asking then lock it up and charge $20 a month for it.
@BenMoore-d1w
April 15, 2025 at 7:54 pm
Healthy holistic Universal AI
@50_Pence
April 15, 2025 at 7:58 pm
the way he walked off at the end says it all. Very worrying indeed
@BenMoore-d1w
April 15, 2025 at 8:14 pm
Ongoing process, work in progress, live and learn, trial and error, experience knowledge and wisdom
@taylerchristensen
April 15, 2025 at 8:38 pm
People have dedicated their lives to developing art styles, fonts, writing styles, etc. You can’t just steal that and market it to the world. The aragance of Sam Altman is baffling. Yes we all agree ai will be amazing for humanity. But companies don’t have the right to avoid IP and copyright infringement by copying 0.1% of a 1000 artists(these numbers are just to paint the picture).
@taylerchristensen
April 15, 2025 at 8:41 pm
Every example of what he says the ai wont do, it does! It can easiy be told to copy writing style, music, art, fonts, etc etc.
@50_Pence
April 15, 2025 at 8:50 pm
Great point but too late. It’s already been done. You see what zuck did with z library?
@OfficialStevenCravis
April 15, 2025 at 8:43 pm
To see the scariest potential uses just watch every season of Black Mirror.
@JLOOV-JLOOV
April 15, 2025 at 9:17 pm
The third AI wave is the loudest one. Any small or tiny thing “made” by the algorithms is amplified to increase the value of the stock. There isn’t a free lunch.
@dennis-qu7bs
April 15, 2025 at 9:26 pm
Sam Altman stole my heart ❤️