Google's CEO Admitted Incomplete Comprehension of Their AI's Functionality

Listen to the Podcast:

You can open Table of Contents show

Sundar Pichai, the CEO of Google, said that there are still some parts of the AI-powered robot Bard that they don’t understand. He talked, for example, about how his model has shown new skills in recent tests, like thinking, planning, and creativity.

“You don’t know how it all works. And yet you’ve given it to the public?” Scott Pelley, who works for CBS, asked Sundar Pichai a question. “Yes. Let me say it like this… I don’t think we fully understand how the human mind works either,” said the head of Google.

Sundar Pichai said that there is a part of their research that they call “the black box” and that they don’t understand. “Sometimes you can’t tell why he said that or why he was wrong. We have some thoughts, and as time goes on, we get better at understanding him. But that’s where things stand now,” he said.

During the chat, they talked about how a Google AI system picked up the Bangladeshi language all on its own. also of his ability to think about the pain people feel or about things like forgiveness.

Are they just using algorithms to put together talks, or is there more to it than that? “I think we need to be humble about this… I think it’s good that some of these tools are being made available so that people can understand what’s going on. And let’s start talking and arguing,” Sundar Pichai replied.

Google Advocates Regulating AI

Google’s CEO said that he can’t sleep because he’s worried that this AI could be “very damaging if implemented incorrectly.” He thinks that society needs to move quickly to make clear rules for how AI can be used in the business. He also asked for rules that make it illegal to hurt others and agreements between governments to make sure that these new technologies are safe for everyone.

Pichai said that it is a mistake for companies to let competitive forces get the best of them. “You don’t want to release a technology like this when it’s very, very powerful, because it doesn’t give society time to adapt,” Pichai told CBS. He told me that the company is trying to find better ways to deal with these systems, which is why new tests have been put off.

But the tech company in Mountain View is one of the leaders in this race. OpenAI is sometimes smarter than Microsoft or OpenAI. His conversational model, Bard, came out in a test form that was much more limited than ChatGPT. But even Google employees have warned, for example, that the AI robot was made without thinking about the fact that it could make up false or dangerous information.

AI will be more Transformative than Electricity

“I’ve always thought that AI is the most advanced science that people are working on. In the conversation with CBS, the CEO of Google said, “Deeper than fire or electricity or anything we’ve done before.”So he doesn’t think people are ready for AI systems with more power.

He said that the speed at which we can think is different from the speed at which this technology is changing. One example of this is Google. In an interview, Pichai talked about it and said that the company is trying to improve its search engine. Google will make a brand-new one that uses AI, just like Microsoft did with Bing.

The New York Times explains that the new search engine is part of the “Magi” project. The goal is for the new tool to be easier to change than the current system. It would have a tool for making images and a robot built right into the search engine.