How Far Is Too Far? | Living In The Age Of AI

Welcome to YouTube Original Stages, once home to Howard Hughes's Spruce Goose assembly hangar, and home to much of the first Iron Man, filmed 12 years ago. Many happy memories here.

And speaking of taking a look back...

Technology. It's advancing faster and taking less time to be widely adopted than ever before, like as in it took roughly 10,000 years to go from writing to printing press, but only about 500 more to get to email.

Now it seems we're at the dawn of a new age, the age of A.I... Artificial Intelligence. 

What does it mean? I don't know. Tons of folks are working on it, right? Most people don't know that much about it, and of course, there's no shortage of data or opinions.

Anyway, I've heard it said that the best way to learn about a subject is to teach it, but to level with ya, I have a wildly incomplete education... Not in my day job, where I've been A.I.-adjacent for over a decade.

Anyway, I figured now would be as good a time as any to catch up on the state of things regarding this emerging phenomenon. My sense of it is it kind of feels like Pandora's box, maybe... ish?

Much of my understanding on this topic has come from sci-fi stories, which usually depict us heading toward Shangri-La or dystopia.

Like most things, I suspect the truth is probably somewhere in the middle. Now, along the way, we'll demystify some common misconceptions about things we thought we understood, but probably don't, terms such as machine learning, algorithms, computer vision and Big Data, they will be conveniently unpacked to help us feel like we know what we're doing, kinda.

By the way, Pandora's box... wasn't a box. It... was a clay jar. How about that? Demystified.

A.I. is teaching the machine, and the machine becoming smart. Each time we create a more powerful technology, we create a bigger lever for changing the world.

It's an extraordinary time, one of unprecedented change and possibility. To help us understand what's happening, this series will look at innovators pushing the boundaries of A.I and how their groundbreaking work is profoundly impacting our lives and the world around us.

In this episode, we'll meet two different visionaries exploring identity, creativity, and collaboration between humans and machines. 

Intelligence used to be the province of only humans, but it no longer is. We don't program the machines. They learn by themselves.

Mark Sagar, CEO, Soul Machines discusses his project with Robert Downey Jr.

My background's always been a mixture of art and science. I ended up doing a PhD in bioengineering, then I ended up in the film industry, working on King Kong to Avatar, simulating faces.

I'd got to a point in my career where I'd been, you know, lucky enough to win a couple of Academy Awards, so I thought, "Okay, what happens if we actually tried to bring those characters to life, that actually you could interact with?"

So "Baby X" is a lifelike simulation of a toddler. She's actually seeing me through the web camera, she's listening through the microphone. Baby X is about exploring the nature of how would we build a digital consciousness, if it's possible? We don't know if it's possible, but we're chipping away at that problem.

"Problem" is an understatement for what Mark's chipping away at. His vision of the future is one where human and machine cooperate, and the best way to achieve that, he thinks, is to make A.I. as life-like as possible. 

Which is why he began where most life begins... a baby... modeled after his own daughter.

So if we start revealing her layers, she's driven by virtual muscles, and the virtual muscles, in turn, are driven by a virtual brain. Now, these are radically simplified models from the real thing, but nevertheless, they're models that we can explore how they work, because we have a real template that exists, the human brain.

So, these are all driven by neural networks.

"Neural network" is a virtual, much simpler version of the human brain. The brain is the most complex system in our body. It's got 85 billion neurons, each of which fire non-stop,
receiving, processing, and sending information.

Baby X's brain is nowhere near as complex, but that's the goal. Instead of neurons, it's got nodes. The more the nodes are exposed to, the more they learn.

What we've learned is it's very hard to build a digital brain, but where we want to go with it is we're trying to build a human-like A.I. which has a flexible intelligence that can relate to people.

Dr. Ayana Howard, Roboticist.

I think the best kind of systems are when humans and A.I. work together. One of the biggest misconceptions of A.I. is that there is a super-intelligent being, or what we call a generalized A.I., that knows all, can do all, smarter than all of us put together. That is a total misconception.

A.I. is built on us. A.I. is mimicking our thought processes. A.I. is basically an emulation of us.
Like visionaries before him, Mark's a dreamer. The current state of his moonshot, however, is a little more earthbound. Today, most avatars are basically glorified customer-service reps. They can answer simple questions and give scripted responses.

Beats dealing with automated phonelines for sure, but it's a far cry from Mark's ultimate vision to create avatars that can actually learn, interpret, and interact with the world around them, like a real human.

Baby X uses a type of A.I. called "object recognition." Basically, it's how a computer sees how it identifies an object, like a spider, or tells the difference between a spider and a duck.

It's something that you and I do naturally but machines, like Baby X, need to learn from scratch, by basically sifting through enormous piles of data to search for patterns, so that eventually, it can drive a car, or pick out a criminal in a crowded photograph, or tell the difference between me and... that guy.

Now she's responding emotionally to me as well, so we've gone all the way down to virtual neurotransmitters, hormones, and so forth, so Baby X has a stress system.

What Mark is working on is known as "affective computing," A.I. that interprets and simulates human emotion.

Rana El Kaliouby, Co-Founder & CEO, Afectiva / Author, Girls Decoded.

I believe that machines are gonna interact with humans just the way we interact with one another, through perception, through conversation. So as A.I. continues to become mainstream, it needs to really understand humans, and so we want to build emotion A.I. that enables machines to have empathy.

Dr. Ayana Howard, Roboticist.

They've showed, for example, older adults who have A.I. aides at their nursing homes, they are happier with a robot that emotes and is social than having no one there. That's really the enhancement of human relationships.

You know, human cooperation is the most powerful force in human history, right? Human cooperation with intelligent machines will define the next era of history. Using a machine which is connected with the rest of the world through the Internet, that can work as a creative, collaborative partner?That's unbelievable., Black Eyed Peas.

I love music. The whole concept of music is collaboration, so if there are some people that see me as a musician, that's awesome. I first became interested in A.I. because A.I. is a very fruitful place to create in. It's a new tool for us. I dream, and make my dreams reality, whether the dream is a song or the dream is an avatar of myself. 

One time, a friend was like, "Well, you can't clone yourself. You can't be in two places at once." That's the promise of the avatar.

The team from Soul Machines is here to create a digital avatar of myself. They had to put me in this huge contraption with these crazy lights.

I wanted to have a digital avatar around the idea of Idatity, and that's the marriage of my data and my identity. Everyone's concerned about, like, identity theft. Meanwhile, everybody's giving away all their data for free on the Internet.

I'm what I like and what I don't like, I'm where I go, I'm who I know. I'm what I search. I am my thumbprint. I am my data. That's who I am. 

In the future, my avatar's gonna be doing all that stuff, because I'm gonna program it. Get entertained through it, get information through it, and you feel like you're having a FaceTime with an intelligent entity.

I'm Will, and I'm happy to meet you. I'm here to bring technology to life, and let's talk about Artificial Intelligence.

The sentences can be divided up into parts so that we can create words and build sentences, like LEGO blocks. It will sound exactly like me. Well, maybe we don't want to have it too accurate. So you don't freak people out, maybe I don't want it accurate.

We are in an intersection of a place that we've never been in society, where people have to determine what's real and what's not.

While Mark jets back to New Zealand to try to create Will's digital doppelganger, Will's left waiting, and  wondering... can Mark pull this off? What does it mean to have a lifelike avatar of you? A digital replicant of yourself? Is that a good idea? How far is too far?

Pedro Domingos, Professor of Computer Science, University of Washington.

We've been collaborating with machines since the dawn of technology. I mean, even today, in some sense, we are all cyborgs already. For example, you use OKCupid to find a date, and then you use Yelp to decide where to go, you know, what restaurant to go to, and then you start driving your car, but there's a GPS system that actually tells you where to go. So the human and the machine decision-making are very tightly interwoven, and I think this will only increase as we go forward.

Human collaboration with intelligent machines... A different musician in a different town with a different approach is giving the same problem a shot.

Gil Weinberg, Professor & Founding Director, Georgia Tech Center for Music Technology discusses his project with Robert Downey Jr.

People are concerned about A.I. replacing humans, and I think it is not only not going to replace humans, it's going to enhance humans.

I'm Gil Weinberg. I'm the founding director of Georgia Tech Center for Music Technology.

In my lab, we are trying to create the new technologies that will explore new ways to be expressive... to be creative...

Shimon, it's a marimba-playing robot. What it does is listen to humans playing, and it can improvise. Shimon is our first robotic musician that has the ability to find patterns, so, machine learning.

Machine learning is the ability to find patterns in data. So, for example, if we feed Shimon Miles Davis, it will try to see what note is he likely to play after what note, and once it finds its patterns, it can start to manipulate it, and I can have the robot playing in a style that maybe is 30% Miles Davis, 30% Bach, 30% Madonna, and 10% my own, and create morphing of music that humans would never create.

Gil's groundbreaking work in artificial creativity and musical expression has been performed by symphonies around the world... but his innovation also caught the attention of another musician... Okay... a guy who unexpectedly pushed Gil beyond enhancing robots to augmenting humans.

I met Jason Barnes about six years ago, when I was just about finishing one phase of developing Shimon, and I was starting to think, "What's next?"

Jason Barnes, Musician.

I got my first drum kit when I was 15, on Christmas, and when I lost my limb, I was 22, so I was kind of used to having two limbs. I started trying to fabricate prosthetics to try and get me back on the kit, which eventually led me to working and collaborating with Georgia Tech.

He told me that he lost his arm, he was devastated, he was depressed, music was his life, and he said, "I saw that you develop robotic musicians. Can you use some of the technology that you have in order to allow me to play again like I used to?"

So that's the prosthetic arm that we built for Jason. When he came to us, he just wanted to be able to use sensors here so he can hold the stick tight or loose.

I suggested "Let's do that, but also, let's have two sticks. One stick can operate with a mind of its own, understanding the music and improvising. One stick can operate based on what you tell it with your muscle, and also, each one of the sticks can play 20 hertz... faster than any humans, and together, they can create polyrhythm, create all kind of textures that humans cannot create."

In some ways, the robotic drum arm allows Jason to play better than he ever has, but it still lacks the true function, or feeling, of a human hand.

This revelation drove Gil to his next innovation... the Skywalker Hand. Inspired by Luke Skywalker from Star Wars, and created in collaboration with Jason, the revolutionary tech brings what was once the realm of sci-fi a little closer to our galaxy.

Currently, most advanced prosthetic hands can't even thumbs-up or flip you the bird. They can only open or grip, using all five fingers at once. Most of the prosthetics that are available on the market nowadays, actually use EMG technology, which stands for "electromyography," and essentially what it does is there are two sensors that make contact with my residual limb, and they pick up electrical signals from the muscles...

So again, when I flex and extend my residual limb, it will open and close the hand, and I can rotate as well, but the problem with EMG is it's a very vague electrical signal, so zero to 100%. It's not very accurate at all.

The Skywalker Hand actually uses ultrasound tech. Ultrasound provides an image, and you can see everything that's going on inside of the arm.

Ultrasound uses high-frequency sound waves to capture live images from inside the body. As Jason flexes his muscles to move each of his missing fingers, ultrasound generates live images that visualize his intention. The A.I. then uses machine learning to predict patterns, letting a man who's lost one of his hands move all five of his fingers individually, even if he's as unpredictable as Keith Moon.

Dr. Ayana Howard, Roboticist.

The work that Gil is doing is really important. Gil comes from a non-engineering background, which means that his technology and the way he thinks about robotics is actually quite different than, say, the way I would think about it, since I come from an engineering background. And the commonality is that we want to design robots to really impact and make a difference in the world.

We were able to create a proof of concept with Jason Barnes. Once we discovered that we can do this with ultrasound, immediately I looked at, "Hey, let's try to help more people."

Manual robotic prosthetics have not been adopted well. Amputees try them, and then they don't continue to use them.

What began as one man's pursuit to innovate music through A.I. and robotics unexpectedly became something much greater. A human body cooperating with a bionic hand is one thing... but is it possible to humanize a machine to the point that it truly seems lifelike? Or is that still sci-fi, and far, far away?

Will's avatar is actually Mark's first go at creating a digital copy of a real person. Wow, that's looking pretty good.

He's not just trying to clone a human, by any stretch, but trying to create an artificial stand-in that's somewhat believable. Still, like most firsts, it's bumpy, and it's a cautious road into the unknown., Black Eyed Peas.

It's awesome where we are in the world of tech. Scary where we are, as well. My mind started thinking, like, "Wait a second here. Why am I doing this? What's the endgame?" Because, eventually, I won't be around, but it would.

Will's endgame is more modest than Mark's: a beefed-up Instagram following, a virtual assistant, anything that might help him expand his creative outlets or free up time for more creative or philanthropic pursuits. 

Okay, so, here we go. That's looking really different. It's gonna be really interesting, because, you know, it's not every day you get confronted with your virtual self. Right. Does he feel that this is like him? If it's not representative of him or if he doesn't think it's authentic, then he won't want to support it.

Ending statement from Sagar.

How far do you go down that path until you can label it a living... a digital living character? This raises some of the deepest questions in science and philosophy, actually, you know, the nature of free will. How do you actually build a character which is truly autonomous? What is free will? What does it take to do that?

Ending statement from Weinberg.

Artificial Intelligence is crucial to the work we are doing, to inspire, to surprise, to push human creativity and abilities to uncharted domains. Unbelievable.

Ending statement from Downey.

Free will... it's something we've been grappling with for thousands of years, from Aristotle to Descartes, and will continue to grapple with for a thousand more. Will we ever be able to make an A.I. that can think on its own? A second, artificial version of me that is truly autonomous? A Robert that can actually think and feel on his own, while this Robert here takes a nap?

Impossible? Well, when you consider what human cooperation has already accomplished... a man on the moon... decoding the human genome... discovering faraway galaxies... I'd put my money on dreamers like Mark and Gil over the "Earth is flat" folks any day. Until then... nap time.

Next episodes...
  • Look at our world today. Look at everything we've created. Artificial Intelligence is gonna be the technology that takes that to the next level.
  • Artificial Intelligence can help us to feed the world's population.
  • The fact that we can find where famine might happen, it's mind-blowing. These are conflict areas, this is an area that we need to look at protecting. Then launch A.I.
  • We are going to release the speed limit on your car. Tim, can you hear me?
  • With A.I., ideas are easy, execution is hard.
  • Even though the limb is synthetic materials, it moves as if it's flesh and bone.
  • You start to think about a world where you can prevent disease before it happens.
  • A.I. can give us that answer that we've been seeking all along... "Are we alone?" Bah!
  • I love the idea that there are passionate people dedicating their time and energy to making these things happen.
  • What excites me the most about where we might be going is having more super-powers... and A.I. is super-powers for our mind.

Transcript from YouTube Original. Edited by Educazio.

Post a Comment