AI is good at seeing patterns, but it’s humans who figure out what to do next, says technologist Priyanka Vergadia. (Recorded at TEDNext 2025 on November 10, 2025)
Join us in person at a TED conference:
Become a TED Member to support our mission:
Subscribe to a TED newsletter:
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.
Watch 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
@ridingboy
March 23, 2026 at 11:20 am
She isn’t backing up her claim with proves at all. Probably nothing what makes us humans human is fundamentally impossible for A.I. in the future. Which is o.k. for me, even exciting. Greetings from a German Artist.
@torieadore1
March 23, 2026 at 11:50 am
You are an artist and you are in favor of AI? Why? How? Currently AI models are stealing from millions of artists everywhere to inform their work. It’s theft. In fact, my daughter is also an artist. She and her fiance have recently set up a non-profit called Feed the Artists so that they can protect their work. I would strongly suggest that as an artist you should consider protecting your work against AI theft. It’s a huge concern.
@Aaaaaa006
March 23, 2026 at 12:08 pm
@torieadore1 I am an artist too and freely give away my work for training AI
@pockets4230
March 23, 2026 at 12:42 pm
What about dancing? Can AI FEEL that?! Human experience is the key here.
@ridingboy
March 23, 2026 at 2:56 pm
@torieadore1 I support universal basic income for everyone. I don’t deny that the idea of intellectual property can serve as a helpful a legal construct in the short term, also to artists, but in the big picture of things, we have always been inspired by an endless mix of what had been created from other artists… culture and the history of art has always worked like that. If an A.I is influenced / informed by my work, so be it. Especially because (and not despite) I am quite self-assured about the uniqueness of my work, I fear not, but welcome the advent of A.I. in the ecosystem.
@ridingboy
March 23, 2026 at 3:20 pm
@pockets4230 Why not, especially taking into account the breathtaking (and still accelerating) pace A.I and robotics are evolving? Robots with A.I. will probably be no less sentient and self aware beings than larger animals in the near future. Some of them will also enjoy dancing while feeling the flow of air on their robot skin. There will be ethics to protect conscious robots.
@Feel_the_ASI
March 23, 2026 at 11:24 am
“we understand things that cannot be quantified”
Why does TED keep bringing on these morons. There’s nothing special about the human brain it’s finding patterns in data and reasoning over its own internal embeddings. AI will be able to do anything the human brain can and better in the future.
@pockets4230
March 23, 2026 at 12:27 pm
What about feelings? Like Ai won’t be able to truly know what it feels like to get tickled or to have fingers chopped off. Aren’t these traits unique to a living organisms nervous system regardless of how much data is acquired?
Not trying to argue as a side point, just discussing.
@ExistentialWolf
March 23, 2026 at 11:53 am
Your lack of understanding of the English language is more an encumbrance for your computer, than the computer’s software correlation weighting. The computer will highlight regular deviations and divergent error correlations, but the scope is obviously limited in inception, and typically in public reception. Listen, we don’t need screens and spreadsheets anymore than we need a crystal ball. The goal in most endevours with human curiosity is enjoyment. I think we’re done here. 🕹⚾
@ExistentialWolf
March 23, 2026 at 11:59 am
😯😯😬😯
@meanderinoranges
March 23, 2026 at 11:57 am
AI slop.
@jamesmitchell6925
March 23, 2026 at 12:27 pm
It doesn’t matter when free and fair elections are at risk because FOLKS DON’T KNOW WHAT TO BELIEVE! Democracy itself is under threat here, people. Wake up! Listen to Bernie sanders and let’s put a moratorium on all new ai so legislators can gain an upper hand on this rapidly growing industry.
@thelonewolf2023
March 23, 2026 at 12:31 pm
Physics numericals. I don’t know why , chat gpt always makes mistakes if a physics problem isn’t direct and requires more than 2 concepts to answer . So I can confidently say that .
@angelic8632002
March 23, 2026 at 12:36 pm
So, who wants to tell her?…
@fburton8
March 23, 2026 at 12:42 pm
Horses. Real life horses.
@ridingboy
March 23, 2026 at 3:05 pm
Yes.
@user-kq2tc5hy5g
March 23, 2026 at 12:45 pm
What AI can never do … joke’s on you
@y0lks
March 23, 2026 at 1:06 pm
some more neoliberl BS smh
@erikkoning8795
March 23, 2026 at 1:12 pm
8:20 if you ask an ai that it will do a great job… We need some better examples of where AI struggles.
@hippityhopotus7475
March 23, 2026 at 1:35 pm
Whelp, AI knows now.
@zaoman69
March 23, 2026 at 2:18 pm
It feels like she’s speaking in a very generic, scripted AI-like way.
@soldierforrester2401
March 23, 2026 at 2:48 pm
I disagree partly, if you wear smart glasses, AI can very much read the emotions in the room.
But AI it a glorified copy&paste machine that can only know what it has been trained on. Current machine learning models are entirely incapable of creativity, inventing something new that has never existed. The combination of training data might be interpreted as new output, but there is a distinct lack of innovation.
@AdityaDodda
March 23, 2026 at 3:31 pm
This talk assumes that AI will never get the signals we get as humans.
@MyMehdiMoments
March 23, 2026 at 7:10 pm
Haha, that part she ignored
@Sayedamahdy
March 23, 2026 at 3:43 pm
❤❤
@pmyou2
March 23, 2026 at 3:55 pm
Actually, no.
The stories you provided list only 1st order analysis by the AI, not 2nd order. AI can do the 2nd order, too, so long as it is asked to do it. AI can detect statistics that show that increased engagement did not increase sales as was the intent for the increased engagement. How did the human identify that it was the wrong audience? Would that data have been available to the AI? I expect so. The proper analysis would have been provided if only the correct prompt was given.
@CptBeton
March 23, 2026 at 5:03 pm
Hmmm, so the person selling AI models tells us we should buy it and use it to be successful and keep in a mind, that a workplace (all workplaces) are safe after further and further training AI models with specific data from yours work?
@betusethu
March 23, 2026 at 5:31 pm
I guess people in these stories are asking wrong questions to AI
@LaughingInChurch
March 23, 2026 at 5:31 pm
Absolutely loved your talk!!!! Congrats!
@LaurenDeeley
March 23, 2026 at 5:31 pm
Absolutely loved your talk!!!! Congrats!
@rob9086
March 23, 2026 at 5:33 pm
How does she say the word “algorithm”?
@sasangaabeywickrama1303
March 23, 2026 at 6:13 pm
All the examples mentioned here are addressable/doable with ai solutions.
@MyMehdiMoments
March 23, 2026 at 7:10 pm
This talk sounds like what people said before humans learned how to fly. The belief was than we won’t fly in a 1000 years, and suddenly we did.
Same counts for a chess robot, we thought AI could never win from humans. Then we thought AI will never become good at language, and then we have it.
Point being? Make more careful statements.
@markshaar5874
March 23, 2026 at 7:34 pm
Probably good insight on the limitations of ai today, but nothing to do with if your job will exist in 10 or 20 years.
@graciousfury
March 23, 2026 at 9:05 pm
What You Know That AI Doesn’t…anything.
@ladyJustis
March 23, 2026 at 9:53 pm
World population control has begun
Most of the population will become irrelevant😮
@ElonMusk567-f2n
March 23, 2026 at 10:24 pm
If you had +124 small potatoes and 98oz of 97oz of ginger root… 34 turmeric and nineOz of v89 juice cooked overnight and covered in low flame… set the alarm to and send a text if you need my help
@osamaelshewy9511
March 24, 2026 at 12:54 am
“The future belongs to humans who work closely with AI” — totally agree. Thanks for this great talk!
@kangkangfoever
March 24, 2026 at 5:09 am
😂😂not great at all actually
@TharinduAsiri-y9k
March 24, 2026 at 3:21 am
🎉
@JungleKing-e5j
March 24, 2026 at 3:21 am
🎉
@Freja-c3o
March 24, 2026 at 5:38 am
Everyone thinks about how they can make money from AI and how they can get ahead in the capitalist race. But I don’t. I only think about how I can destroy the entire capitalist system.
And I have wild fantasies about money becoming worthless.
@samueladams1565
March 24, 2026 at 8:59 am
You’re late, money is already worthless, power is the only currency
@samueladams1565
March 24, 2026 at 9:03 am
Can AI ……lie
@TraumaER
March 24, 2026 at 10:26 am
This speech won’t age well 😅
@6rumpybear
March 24, 2026 at 12:00 pm
ted is a waste of time nowadays, sad
@okayssorta
March 24, 2026 at 1:14 pm
lol, that’s a pretty naive approach. “Being human”, tell this to your capitalist corporate overlord, and watch them laugh so maniacally that you won’t need an AI to understand that you are fired, and are replaced by an API for the latest model.
@FiveRomeoCharlie
March 24, 2026 at 1:49 pm
This video will not age well. Anyone that says that humans can do something that AI can never do simply doesn’t understand the potential of AI. There is nothing special about computers made of meat.
@christiannider1353
March 24, 2026 at 2:36 pm
“Dont compete with AI but work with it while staying irreplaceably human” – Love this!
@landdcollection
March 24, 2026 at 11:59 pm
The people in her example sound like people who don’t know how to use AI, frankly. AI can do all of the “human only” things she says — I use it to do that at work daily. Signed, an executive whose job could very well be taken by AI
@mitch14311
March 25, 2026 at 1:13 am
More Ai propaganda promising AGI, anything to keep the bubble growing. If you always ask “what happens when Ai takes your job” you’ll never ask “can Ai even take my job” fear tactics to sell a product that isn’t anywhere near that reality
@Hips4444
March 25, 2026 at 1:29 am
This seems to just a be a collection bad data collection, or poor/limited prompting stories
@darraghogrady3350
March 25, 2026 at 6:44 pm
Love this! Critical thinking applied