“We’re working together with Microsoft to build next-generation supercomputers,” says OpenAI co-founder Greg Brockman. “The real goal of OpenAI and what we’re trying to accomplish is to build what we call artificial general intelligence. They’re trying to build a computer system that is as capable as a human at being able to master a domain of study and being able to master more fields of study than any one human can. We think whoever builds artificial general intelligence will be the number one most valuable company in the world by a huge margin.”
Greg Brockman, OpenAI co-founder and chairman, discusses Microsoft’s $1 billion investment in OpenAI and how whoever invents artificial general intelligence first will become the most valuable company in the world. Brockman was interviewed by Bloomberg Technology:
OpenAI Working With Microsoft To Build AI That Will Change The World
We’re working together with Microsoft to build next-generation supercomputers. The real goal of OpenAI and what we’re trying to accomplish is to build what we call artificial general intelligence. They’re trying to build a computer system that is as capable as a human at being able to master a domain of study and being able to master more fields of study than any one human can. If we succeed the kind of thing that we want to be able to do is, for example, build a computer system that can solve medicine better than humans can. If you think about how humans solve medicine today we do it by increased specialization.
I have a friend who’s going through medical procedures right now where he talks to a first doctor who does an ultrasound but can’t read it so he has to go to a different doctor who doesn’t really have context as to what’s going on. This is not a problem that we can solve by increasing the amount of knowledge that humans have to learn. There’s only so much we can fit in our minds. What we really need are tools that are capable of helping us solve these problems. That’s the kind of thing that we want to apply general intelligence to.
Our goal is to distribute the economic benefits of artificial general intelligence. You can imagine a general intelligence system that can generate huge amounts of value. If you look at the top ten most valuable companies in the world, seven of them are technology companies. We think whoever builds artificial general intelligence will be the number one by a huge margin. It’s really important that those benefits go to everyone rather than being locked up in one place.
Building Powerful Safe and Secure AI Technology
There’s a second part which is it’s really important that you keep these systems safe and secure and that you build them with ethics in the forefront. That’s something that both we and Microsoft are very aligned on doing from the beginning. What it really boils down to is that AI technology is becoming very powerful. That means that there’s both these amazing benefits and these amazing applications. Imagine a personalized tutor that can really understand you that is available for free to every person on the planet. That’s the kind of thing we should be able to build with the kind of systems that we want to create.
You also have to ask the questions of what are the risks. How can they be misused? Today, we already see AI technology, for example, deepfakes, that already has bad implications in the world. How do we maximize those benefits and mitigate the downsides? That’s our goal. Our goal is to push the technology forward and make sure that we’re capturing those benefits while making sure everyone benefits from them. But we also want to make sure that we keep it safe and secure to mitigate the downsides.
AI Computational Power Growing 5 Times Faster Than Moore’s Law
The timelines (of where AI will take us) are always really hard to predict. One story I really like thinking about is just looking, for example, at previous technological innovations. In 1878, Thomas Edison announced that he was going to create the incandescent lamp and gas securities in England fell. So British Parliament put together a commission of distinguished experts who came out to Menlo Park. They checked everything out. They said this isn’t going to work and one year later he shipped. I think that we’re in a similar sort of place here where it’s hard to predict what the future will be like.
We’re in this exponential right now where the computational power that we’re using is growing five times faster than Moore’s Law. What we do know is every year we’re going to have unprecedented AItechnologies. We’ve been doing that for seven years and OpenAI has been doing it for three. This year we have systems that can understand and generate text. I think five years from now we should expect that we will have systems that you can really have meaningful conversations with. I think that we should see within a bunch of different domains, a lot of systems that can work with humans to augment what they can do much further than anything we can imagine today.