AI is learning to create yourself

But there is another crucial observation here. Intelligence has never been an end point for evolution, something to aim for. Instead, it has emerged in many different forms from countless tiny solutions to challenges that allow living things to survive and face future challenges. Intelligence is the current culmination in a continuous and open process. In this sense, evolution is quite different from the algorithms of the way people typically think of them — as means to an end.

It is this open-ended, glimpse into the seemingly aimless sequence of challenges generated by POETs that Clune and others believe could lead to new types of AI. For decades AI researchers have tried to build algorithms to mimic human intelligence, but the real breakthrough may come from building algorithms that try to mimic the open-ended solving of evolutionary problems — and to sit back and watch what emerges. .

Researchers are already using machine learning on their own, training them to find solutions to some of the field’s toughest problems, such as how to make machines that can learn more than one task at a time or deal with situations they don’t have. not encountered before. Some think now that taking this approach and running with it might be the best path for general artificial intelligence. “We could start an algorithm that initially doesn’t have a lot of intelligence in it, and look at bootstrap itself until potentially at AGI,” says Clune.

The truth is that for now, AGI remains a fantasy. But it’s largely because no one knows how to do it. Advances in AI are partial and man-made, with progress typically involving modifications to existing techniques or algorithms, giving incremental jumps in performance or accuracy. Clune characterizes these efforts as attempts to discover the building blocks for artificial intelligence without knowing what you are looking for or how many blocks you need. And that’s just the beginning. “At some point, we have to take on the Herculean mission of putting them all together,” he says.

Asking AI to find and assemble those blocks for us is a paradigm shift. It is said that we want to create a smart machine, but we don’t care how it might look – just give us what works.

Although AGI is never implemented, the self-taught approach can still change what type of AI are created. The world needs more than a good Go player, says Clune. For him, creating a supersmart machine means building a system that invents its own challenges, solves them, and then invents new ones. POET has a small vision of this in action. Clune imagines a machine that teaches a bot to walk, then play to jump, then maybe play to Go. “So maybe he learns math puzzles and starts inventing his own challenges,” he says. “The system is constantly innovating, and the sky is the limit in terms of where it could go.”

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button