Does Generative AI represent an evolutionary next step in a 70-year-old idea—dating from the Alan Turing era—or a phase-change future shock? How we organize our companies, investments, educational systems, governments and societies in the next two decades depends on our answers.
The final panel at last month’s Forbes Global CEO Conference examined this vital question.
Antoine Blondeau, cofounder and managing partner, Alpha Intelligence Capital
Phase change. It’s the difference between monkeys and humans, right? Monkeys, just like us, are very good at perceiving. They see, identify characters, pick a threat. We’re good at that, too. They may be even better than us. Monkeys are also good at optimizing. The very fact that they thrive, survive, adapt, they’re good at optimizing. But cognition is where it’s different. Monkeys are not good at cognition. We are very good at cognition. We can build our own models of the world. Gen AI is moving AI from perception and optimization to cognition. Machines can now build their own cognition model without having to learn the way we do. That’s remarkable. And unexpected. It’s a case where, maybe for the first time in a long time, maybe forever, technology is completely leapfrogging expectations.
Meng Ru Kuok, group CEO and founder, Caldecott Music Group
Definitely phase change. The speed of this unexpected technology shift has changed our whole perspective on the opportunities at hand. Someone asked me recently, ‘Why is modern music bad?’ Or, ‘I like Fleetwood Mac, I like BB King, why is today’s music different to that kind of stuff, or music from 20, 30, 40 years ago?’ And the answer is, we’re different. We aren’t humans anymore in some ways. We’re all augmented because of our phones. Therefore our attention span is different. What we need to know, or don’t know, has changed. I don’t need a map to get around Singapore. I just know how to get there.
Rohan Narayana Murty, founder and chief technology officer, Soroco
I’ll give you a more nuanced answer. I don’t think Gen AI is cognition. It looks like cognition. It simulates certain parts of human cognition. But what Gen AI really does is changes the relationship of machines to words. What we see at Soroco is how teams get work done in companies—it is actually about words. Because there are patterns to work, there are work patterns that can be described with words. And therefore, suddenly now you can say, I know how to optimize work. Pretty much the work consultants do. It all now theoretically is to be done by AI. All these seem quite crazy to me.
Eduardo Saverin, cofounder and co-CEO, B Capital
I can think about this in several ways. One of my favorite pastimes as a child was playing chess. Back in the late 1990s, we were talking about Deep Blue playing a six-game series with chess grandmaster Garry Kasparov and succeeding. Deep Blue was an early version of a computer being able to outsmart the top player in the world by knowing all the great historical chess moves and using them in a very deep reasoning-style game. Years later you had Google’s Alpha Zero. Alpha Zero simply derived and got to the solutions of how to play chess without being trained by historical chess moves. It simply observed chess games and learned. Now Gen AI is another massive acceleration of a very long pattern of leveraging technology toolsets. The phase change will happen—and we’re not very far from it—when computers can effectively program themselves.
Read the full article here
