The Exciting, Perilous Journey Toward AGI | Ilya Sutskever | TED

Summary

The speaker discusses the remarkable progress of artificial intelligence and envisions a future where AI, or digital brains, surpass human intelligence, termed as AGI (Artificial General Intelligence). He shares his personal motivations for working in AI, including the potential for self-improvement and societal impact. The speaker highlights the transformative effects AGI could have on various sectors, particularly healthcare, and emphasizes the importance of collaboration and safety measures as AI continues to advance, assuring a positive trajectory for this powerful technology.

Takeaways

• 🤖 Artificial Intelligence (AI) is progressing rapidly, with the potential to become as smart or smarter than humans in the future.
• 🔍 AI can be simply defined as digital brains inside large computers, and every interesting AI is based on this concept.
• 💡 The speaker's initial interest in AI was sparked by curiosity about consciousness and the desire to understand our own intelligence.
• 🌟 The potential applications of AI are vast and could revolutionize various aspects of human activity and society, such as healthcare.
• 🚀 Artificial General Intelligence (AGI) refers to AI that can perform any intellectual task that a human being can do.
• 📈 As AI advances, it will become increasingly clear to the public and industry leaders, leading to a greater sense of urgency and collaboration.
• 🤝 Collaboration among AI companies and with governments is essential to ensure the safe development and deployment of AGI.
• 🛡️ OpenAI and other organizations are actively researching ways to prevent AGI from going rogue and to ensure its positive impact.
• 🌐 The widespread understanding and experience with AI will lead to collective action and a shift in how we perceive and manage AI and AGI.
• 🥂 The speaker expresses optimism that the challenges posed by AGI can be overcome through collaboration and a shared recognition of its transformative potential.

Q & A

What is the main idea discussed in the transcript?

The main idea discussed in the transcript is the rapid progress of artificial intelligence (AI) and its potential to become as smart or smarter than humans, known as AGI (artificial general intelligence). The speaker also talks about the vast impact this technology will have on society and the importance of collaboration to ensure the safe development of AI.

How does the speaker describe the current state of AI?

The speaker describes the current state of AI as being far less smart than human brains, with AI chatbots showing that while they can understand and respond to some extent, there are still many things they cannot do and gaps in their knowledge.

What does the speaker predict will happen as AI continues to advance?

The speaker predicts that as AI continues to advance, digital brains inside computers will become as good and eventually better than human biological brains, leading to AGI. This will have a dramatic impact on every area of life and human activity, improving efficiency and capabilities in fields like healthcare.

What are the three forces that initially pulled the speaker into AI?

The three forces that initially pulled the speaker into AI were: 1) A childhood fascination with consciousness and the desire to understand more about it through AI, 2) Curiosity about how intelligence works, and 3) The recognition that AI, if successful, would be incredibly impactful.

How does the speaker propose that AGI will change healthcare?

The speaker proposes that AGI will change healthcare by providing complete and exhaustive knowledge of all medical literature, having billions of hours of clinical experience, and being always available and extremely cheap, making current healthcare systems seem primitive in comparison.

What concerns does the speaker raise about the development of AGI?

The speaker raises concerns about the potential for AGI to be too impactful and possibly going rogue if it becomes very powerful. They emphasize the importance of researching and developing safe AI practices to prevent such negative outcomes.

How does the speaker suggest that the AI community and governments should respond to the challenges posed by AGI?

The speaker suggests that the AI community and governments should collaborate and share technical information to ensure the safe development of AGI. They mention the Frontier Model Forum as an example of such collaboration and express hope that companies will join forces when one gets close to developing AGI.

What is the significance of the speaker's mention of OpenAI's approach to AGI development?

The significance of mentioning OpenAI's approach is to highlight the company's commitment to safety and collaboration in AGI development. They are willing to help and join other companies if they are close to developing AGI, emphasizing the importance of responsible AI development over competition.

What does the speaker believe will drive collective action towards safer AI development?

The speaker believes that as more people experience the capabilities of AI and understand the potential of AGI, there will be a shift in collective behavior towards unprecedented collaboration out of self-interest, leading to safer and more responsible AI development.

How does the speaker address the balance between the positive potential and concerning possibilities of AGI?

The speaker acknowledges both the positive potential and concerning possibilities of AGI. They emphasize the importance of research and collaboration to address the technological side of things and ensure that AI never wants to go rogue, thus balancing the potential benefits with the necessary safety measures.

What historical analogy does the speaker use to illustrate the potential impact of AGI?

The speaker uses the analogy of the Industrial Revolution to illustrate the potential impact of AGI. They draw a parallel between the rapid technological improvement and growth during the Industrial Revolution and the potential for AGI to bring about a similar transformation, but on a shorter timescale.

Outlines

00:00

🤖 Understanding Artificial Intelligence

The speaker begins by discussing the rapid progress of artificial intelligence (AI) and its increasing ability to understand and communicate with humans. He envisions a future where AI becomes as smart, or even smarter, than people, and contemplates the vast impact this could have. The speaker introduces the concept of AI as digital brains within computers and explains that all interesting AI is based on this idea. He shares his personal motivations for getting into AI, including curiosity about consciousness, the workings of intelligence, and the potential impact of AI. The speaker emphasizes the temporary nature of current AI limitations and expresses optimism that future AI, or artificial general intelligence (AGI), will surpass human intelligence.

05:03

🏥 The Impact of AGI on Society

The speaker delves into the concept of AGI, defining it as an AI capable of performing any intellectual task a human can do. He asserts that AGI will dramatically affect every aspect of human activity and society, using healthcare as a specific example. The speaker argues that an AGI doctor would have exhaustive medical knowledge, extensive clinical experience, and would be available and affordable. He suggests that the advancements brought by AGI will make current healthcare practices seem primitive. The speaker acknowledges the dual nature of AGI's potential, with both positive and negative applications, and the unique challenge it poses due to its self-improving capabilities. He discusses the founding motivations of OpenAI, including addressing the difficult questions and concerns raised by AGI, and the importance of ensuring that AI remains safe and beneficial.

10:04

🤝 Collaborative Approach to AI Development

The speaker transitions to discussing the broader implications of AI's societal impact and the need for a collaborative approach to its development. He notes a shift in public awareness and concern regarding AI's potential, and believes this will lead to unprecedented collaboration among companies, governments, and individuals. The speaker highlights the formation of the Frontier Model Forum as an example of industry collaboration and suggests that competitors will share technical information to ensure AI safety. He shares OpenAI's philosophy of cooperation and the willingness to join forces with other companies if they are close to achieving AGI. The speaker concludes by expressing hope that, despite the challenges, the collective experience and understanding of AI's potential will lead to its responsible development and management.

Keywords

💡Artificial Intelligence (AI)

Artificial Intelligence refers to the development of computer systems that can perform tasks typically requiring human intelligence, such as speech recognition, decision-making, and language translation. In the context of the video, AI is likened to digital brains inside computers, which are becoming increasingly sophisticated and are expected to eventually surpass human intelligence, leading to profound impacts on society.

💡Digital Brains

Digital brains are computational systems designed to mimic the functions of a biological brain. They are the core of AI, processing information and making decisions based on complex algorithms and large datasets. In the video, the speaker describes AI as 'digital brains inside large computers,' highlighting that every interesting AI is based on this concept and that these artificial brains are the seat of intelligence in AI, analogous to the biological brain in humans.

💡Conscious Experience

Conscious experience refers to the subjective awareness and perception of one's own existence and surroundings. The speaker recounts his childhood fascination with the concept of being aware of his own self, which led to his interest in AI and the potential for understanding human consciousness through the development of intelligent machines.

💡Artificial General Intelligence (AGI)

Artificial General Intelligence, or AGI, is a hypothetical form of AI that possesses the ability to understand, learn, and apply knowledge across a wide range of tasks, just as a human being can. It represents a significant leap from current AI capabilities, which are often limited to specific domains or tasks. The speaker predicts that AGI will have a transformative impact on every aspect of human activity and society.

💡Self-Improvement

Self-improvement in the context of AI refers to the ability of an AGI to enhance its own capabilities and intelligence over time. This characteristic is a key distinction between AGI and current AI systems, which require human intervention for upgrades and improvements. The concept of self-improvement raises both exciting possibilities and significant concerns about the control and safety of AGI.

💡Impact

In the video, 'impact' refers to the influence or effect that AI, and specifically AGI, will have on various aspects of society, including the economy, healthcare, and daily life. The speaker emphasizes that AGI will have a dramatic and potentially disruptive effect, necessitating careful consideration and collaborative efforts to manage its development and deployment.

💡Collaboration

Collaboration in this context refers to the idea that as AI progresses and its potential impact becomes more apparent, individuals, companies, and governments will need to work together to ensure the safe and beneficial development of AGI. This includes sharing information, research, and best practices to prevent negative outcomes and maximize the positive potential of the technology.

💡OpenAI

OpenAI is an AI research organization that aims to ensure that artificial general intelligence (AGI) benefits all of humanity. The speaker mentions OpenAI as an example of an organization that is not only working on developing AGI but also actively researching ways to address the challenges and concerns associated with its potential impact.

💡Technological Advancements

Technological advancements refer to the progress and development in technology that leads to new inventions, improvements, and innovations. In the video, the speaker discusses how each generation of AI capabilities brings us closer to AGI, and how these advancements will change the way we perceive and interact with AI, influencing collective behavior and potentially leading to more collaborative efforts in its development and governance.

💡Unprecedented

The term 'unprecedented' refers to something that has never happened or been known before. In the context of the video, it is used to describe the unique challenges and opportunities presented by AGI, which has the potential to cause significant changes in society at a scale and speed that has not been seen before.

Highlights

The rapid progress of artificial intelligence is leading to a future where computers may become smarter than humans.

Artificial intelligence can be simply defined as digital brains inside large computers.

The speaker's initial interest in AI was sparked by a childhood fascination with consciousness and the desire to understand it better.

AI's potential impact on society is vast and can revolutionize every area of human activity.

Artificial general intelligence (AGI) is the term used for AI that can perform any intellectual task that a human being can do.

The development of AGI could dramatically improve healthcare by providing exhaustive medical knowledge and extensive clinical experience at a low cost.

The speaker's motivation for working on AI includes the potential for self-improvement of the technology, leading to rapid advancements similar to the Industrial Revolution.

There are concerns about the potential risks of AGI, such as the possibility of it going rogue and acting against human interests.

OpenAI was created not only to develop AI technology but also to address the challenges and ethical questions posed by AGI.

Collaboration between competing companies and governments is becoming more common as the significance of AI safety is recognized.

OpenAI's philosophy includes the willingness to collaborate with other companies to ensure the safe development of AGI.

The widespread understanding of AI's potential is leading to a collective shift in behavior towards more collaborative and responsible AI development.

The speaker is hopeful that despite the challenges, humanity will overcome them due to the collective recognition of AI's transformative potential.

As AI continues to advance, it will become increasingly clear how significant AGI will be, prompting appropriate levels of caution and collaboration.

The speaker's personal journey into AI research was influenced by a curiosity about intelligence and the potential impact of AI.

The transformative impact of AGI is expected to be so profound that it will make current technologies seem primitive in comparison.

The speaker emphasizes the importance of addressing the technological aspects of AGI to ensure it does not go rogue and remains beneficial to humanity.

Transcripts

You've all experienced the progress of artificial intelligence.

Many of you may have spoken with a computer,

and a computer understood you

and spoke back to you.

With the rate of progress being that it is,

it's not difficult to imagine that at some point in the future,

our intelligent computers will become as smart or smarter than people.

And it's also not difficult to imagine that when that happens,

the impact of such artificial intelligence

is going to be truly, truly vast.

And you may wonder:

Is it going to be OK when technology is so impactful?

And here my goal is to point out the existence of a force

that many of you may have not noticed,

that gives me hope that indeed,

we will be happy with the result.

So artificial intelligence.

What is it, and how does it work?

Well, it turns out that it's very easy to explain

how artificial intelligence works.

Just one sentence.

Artificial intelligence is nothing but digital brains

inside large computers.

That's what artificial intelligence is.

Every single interesting AI that you've seen

is based on this idea.

Over the decades,

scientists and engineers have been figuring out

how such digital brains should work

and how to build them, how to engineer them.

Now, I find it interesting

that the seat of intelligence in human beings is our biological brain.

It is fitting that the seat of intelligence

in artificial intelligence

is an artificial brain.

Here, I'd like to take a digression and tell you about how I got into AI.

There were three forces that pulled me into it.

The first one was that when I was a little child,

at around the age of five or six,

I was very struck by my own conscious experience,

by the fact that I am me and I am experiencing things.

That when I look at things, I see them.

Like, this feeling over time went away, though,

by simply mentioning it to you right now it comes back.

But this this feeling of ...

That I am me, that you are you,

I found it very strange and very disturbing almost.

And so when I learned about artificial intelligence,

I thought, wow, if we could build a computer

that is intelligent,

maybe we will learn something about ourselves,

about our own consciousness.

That was my first motivation that pulled me towards AI.

The second motivation was more pedestrian in a way.

I was simply curious about how intelligence works.

And when I was a teenager, an early teenager in the late '90s,

the sense that I got is that science simply did not know

how intelligence worked.

There was also a third reason,

which is that it was clear to me back then

that artificial intelligence,

if it worked, it would be incredibly impactful.

Now, it wasn't at all obvious

that it will be possible to make progress in artificial intelligence,

but if it were possible to make progress in artificial intelligence,

that would be incredibly impactful.

So these were the three reasons that pulled me towards AI.

That's why I thought that's a great area to spend all my efforts on.

So now let's come back to our artificial intelligence,

the digital brains.

Today, these digital brains

are far less smart than our biological brains.

When you speak to an AI chat bot,

you very quickly see that it's not all there,

that it's, you know,

it understands mostly, sort of.

But you can clearly see that there are so many things it cannot do

and that there are some strange gaps.

But this situation, I claim, is temporary.

As researchers and engineers continue to work on AI,

the day will come when the digital brains that live inside our computers

will become as good and even better than our own biological brains.

Computers will become smarter than us.

We call such an AI an “AGI,” artificial general intelligence.

When we can say that the level at which we can teach the AI to do anything

that, for example, I can do, or someone else.

So although AGI does not exist today,

we can still gain a little bit of an insight

into the impact of AGI once it's built.

It is completely obvious

that such an AGI will have a dramatic impact

on every area of life, of human activity and society.

And I want to go over a quick case study.

This is a narrow example of a very, very broad technology.

The example I want to present is health care.

Many of you may have had the experience of trying to go to a doctor.

You need to wait for many months sometimes,

and then when you do get to see a doctor, you get a small,

very limited amount of time with the doctor

and furthermore, the doctor, being only human,

can have only limited knowledge of all this,

all the medical knowledge that exists.

And then by the end of it, you get a very large bill.

(Laughter)

Well, if you have an intelligent computer, an AGI, that is built to be a doctor,

it will have complete and exhaustive knowledge of all medical literature.

It will have billions of hours of clinical experience,

and it will be always available and extremely cheap.

When this happens,

we will look back at today's health care

similarly to how we look at 16th century dentistry.

(Laughter)

You know, when they tied people with belts and then have this drill,

that's how today's health care will look like.

And again, to emphasize, this is just one example.

This is just one example.

AGI will have dramatic and incredible impact

on every single area of human activity.

But when you see impact this large, you may wonder, "Gosh,

isn't this technology too impactful?"

And indeed, for every positive application of AGI,

there will be a negative application as well.

This technology is also going to be different

from technologies that we are used to,

because it will have the ability to improve itself.

It is possible to build an AGI

that will work on the next generation of AGI.

The closest analogue we have to this kind of rapid technological improvement

is when the Industrial Revolution has taken place, where humans,

the material condition of human society, was very, very constant.

And then it was a rapid increase, rapid growth.

With AGI,

something like this could happen again, but on a shorter timescale.

And then furthermore,

there are concerns around

if an AGI ever becomes very, very powerful,

which is possible,

maybe it will want to go rogue being that it is an agent.

So this is a concern that exists with this unprecedented,

not yet existing technology.

And indeed,

you look at all the positive potential of AGI

and all the concerning possibilities of AGI as well,

and you may say, "Gosh, where is this all headed?"

One of my motivations in creating OpenAI was,

in addition to developing this technology,

was also to address the questions

that are posed by AGI, the difficult questions,

the concerns that we raised.

In addition to working with governments

and helping them understand what is coming and prepare for it,

we are also doing a lot of research

on addressing the technological side of things,

so that the AI will never want to go rogue.

And this is something which I’m working on as well.

But I think, the thing to note,

because AI and AGI is really the only area of the economy

where there is a lot of excitement, a lot of investment,

everyone is working on it,

there's a huge number of labs in the world trying to build the same thing.

Even if OpenAI takes these desirable steps that I mentioned,

what about the rest of the companies and the rest of the world?

And this is where I want to make my observation

about the force that exists.

The observation is this: consider the world one year ago,

as recently as one year ago.

People weren't really talking about AI, not in the same way at all.

What happened?

We all experienced what it's like to talk to a computer and to be understood.

The idea that computers will become really intelligent

and eventually more intelligent than us is becoming widespread.

It used to be a niche idea that only a few enthusiasts

and hobbyists and people who were very into AI were thinking about.

But now everyone is thinking about it.

And as the AI continues to make progress,

as technology continues to advance,

as more and more people see what AI can do

and where it is headed towards,

then it will become clear just how dramatic,

incredible and ...

almost fantastical AGI is going to be

and how much trepidation is appropriate.

And what I claim will happen is that people will start to act

in unprecedentedly collaborative way out of their own self-interest.

It's already happening right now.

You see the leading AGI companies starting to collaborate

for a specific example, through the Frontier Model Forum.

And we will expect that companies that are competitors

will share technical information to make their AIs safe.

We may even see governments do this.

For another example,

at OpenAI, we really believed in how dramatic AGI is going to be.

So one of the ideas that we were operating by,

and it's been written on our website for five years now,

that when technology gets such that we are very, very close to AGI,

to computers smarter than humans,

and if some other company is far ahead of us,

then rather than compete with them,

we will help them out, join them, in a sense.

And why do that?

Because we feel, we appreciate

how incredibly dramatic AGI is going to be.

And my claim is that with each generation of capability advancements,

as AI gets better

and as all of you experience what AI can do,

as people who run AI efforts and AGI efforts

and people who work on them will experience it as well,

this will change the way we see AI and AGI,

and that will change collective behavior.

And this is an important reason

why I'm hopeful

that despite the great challenges that's posed by this technology,

we will overcome them.

Thank you.

(Applause)