How do we learn to work with intelligent machines? | Matt Beane

Share it with your friends Like

Thanks! Share it with your friends!


The path to skill around the globe has been the same for thousands of years: train under an expert and take on small, easy tasks before progressing to riskier, harder ones. But right now, we’re handling AI in a way that blocks that path — and sacrificing learning in our quest for productivity, says organizational ethnographer Matt Beane. What can be done? Beane shares a vision that flips the current story into one of distributed, machine-enhanced mentorship that takes full advantage of AI’s amazing capabilities while enhancing our skills at the same time.

Check out more TED Talks:

The TED Talks channel features the best talks and performances from the TED Conference, where the world’s leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design — plus science, business, global issues, the arts and more.

Follow TED on Twitter:
Like TED on Facebook:

Subscribe to our channel:


Daichi Hirata says:

The point is “surgeon’s skill vs patients’ benefits.”
I think patients’ benefits are more important so we gotta figure out how new surgeons learn the skills without risking the patients’ body function.

Ivan Liao says:

Talk is 8 minutes too long for something that boils down to "It's important to keep humans in the loop of work from surgery to bomb disposal" and the random little bit added at the end about VR and training.

sctr1235 says:

A machine can know no more than it is programmed to know.

Danny Bou-Maroun says:

I'm sorry, but why is it important to learn something that an AI is already doing better?

elvinperia says:

Now I realized why we were not allowed to use our calculators during our Math exams in high school.

Juni Sulli says:

An A.I. that doesn't know how to work with a human being isn't very intelligent is it?
He is assuming a machine with built in precision guides and auto mechanics is more intelligent because it is more precise therefore it is more intelligent than a human. Considering that the video is using the medical field and surgery to project its conclusion upon us, let's put those same surgical machines in a triage on a battlefield all the while the machine has to completely boot re boot its systems as the power is lost, and what happens? The machine stops being effective at all.
Try again Professor I think that machine just gave you the middle finger!

the ninja club says:

what happens at 4:53

Manuficient says:

This is exactly what they're doing at Impruver Technologies. Check them out at

Akash Lobo says:

Disliked. If robots are very apparently doing the job better than humans, why learn to do the job at all? He completely ignores that paradox. Learn something else that robots cannot do for the moment and when AI can do everything better than us then…. I don't know.

Lorendrawn says:

Yeah this is why I'm studying to be a counselor in my lunch breaks at work.

Invox says:

Watch at x1.5 speed.

Jesska says:

Try to apply his logic for the trains in 1800, cars in 1900, improved agricultural techniques etc. and combined with your knowledge of exponential advancement you will see very obvious fault in this speaker's logic.
For example: There's a simulation for surgeons that's an exponential technology. In VR. You can watch and train to operate on nearly infinite number of patients

There's things like Khan Academy that far exceed whatever capability random math teacher from my school and university had.

This guy lacks fundamental understanding of exponential technology.

clarebelz bellinger says:

well done you – if practicals are taken away in everything, eventually will become defunct – it only takes an emp blast to knock out those robots, and if doctors only know theory, then how would they save lives?! The hubris surrounding technology on ted is astounding at times. Those who lived before it had to memorise procedures… clever or what?

Aykut says:

If happiness is the meaning of life & hence the objective, then a life with less work for a growing population wont fulfil that. Even if ppl ended up receiving universal basic income, they will get bored of leisure. But because theres a conflict of interest between such ppl, with those who believe they will get better results with AI – this is where governments & law enforcement are supposed to step in.

I'd like to see a worldwide survey take place with 1 of 2 multiple choice answers to choose from. Do you have a dream job? Do you not know the career you want to have? I feel as if the 2nd option will receive the most answers & hence AI could be allowed to win

Programmer Parker says:

Sacha Baron Cohen?

TAR ICO says:

Soon, on the IQ scale, we'll cross: Americans in free fall, the machines rising above us. And yes folks: they will look down on us as if we were "special," them ever-so-slightly doing the smh thing, tsking.

Rachel Green says:

That's the kind of future I dream of too.

Fandyus says:

A white male on TED in 2019??? Astounding.

Luis Cortes says:

By following instructions…

Iron Maiden says:

We have already sealed our fates. We stepped over the line when we decided to invent AI. When the Nano bots learn how to repair themselves and reinvent new bots. We are ghosts.

Epic Gamer Times says:

Wait, this show is called Ted Talks, but when will Ted talk?

Sonam Tobgay says:

It's Sacha Baron Cohen everyone. He is Matt Beane by day Borat by night.

Eoin Heffernan says:

Sad to see wonderfully intelligent people reduced unwittingly in this presentation format to lobbyists for broadscale business plans masquerading as simple munificient 'progression'. McKenzie figures predict dictatorship and dominance. They are not simply large percentages of tech dominance in discrete sectors. As the speaker says, today's problems demand we do better, let's first though, quite publicly and clearly define the goals by which better is calibrated. In the corperate realm where AI resides, better means growth. Growth is measured in Fiat currency. Let's finish our sentences where AI is concerned. The most broadly distributed 'benefit' appears to start with the words "universal basic…". All animals are created equal…

____________ says:

Scary how these paranoid trust fund babies would rather work with intelligent robots and undermine the human brains potential and by the looks of it especially his own brains potential is just seething away while hes lecturing about robots –

Susan Nigh Charthaigh says:

I wish there was more communication between people like him and another TED talker who described a process he and his team were working on. They had created a complete chem lab within virtual reality. It followed all the rules of chemistry such that people could create a nuclear bomb and set it off in VR without any harm to themselves. Better than that, they could rewind and slow mo every bit of it including the destruction it did at various intervals from the blast! Something like that is exactly what Matt's team needs. A virtual reality of the human body where they can practice on the procedures. When they move wrong, the VR would produce the exact reaction it would make. It would never hurt a human to learn this way either!


in my opinion, his examples are somewhat far away from the AI in a conventional sense. Apart from that, I agree with his idea.

Write a comment


Area 51