The hype around DeepMind’s new AI model misses what’s actually cool about it

“Nature is trying to tell us something here, which is that this doesn’t really work, but the field relies so much on its own press clippings that it can’t see it,” he adds.

De Fritas’ deep-minded colleagues, Jackie Kay and Scott Reed, who worked with him on the ghetto, were more cautious when I asked him directly about his claims. When asked about whether Geto was going to AGI, he said he would not be drawn. “I don’t think it’s really possible to predict with these kinds of things. I try to avoid it. It’s like predicting the stock market, “said Kay.

Reed said the question was difficult. “I think most machine learning people will avoid answering questions carefully. It’s very difficult to predict, but, you know, hopefully we’ll get there someday. “

In a way, the fact that Deepmind called Geto a “normalist” may have made him a victim of the AI ​​sector’s hype surrounding AGI. Today’s AI systems are called “compressed” AI, which means they can only perform a specific, restricted set of tasks, such as generating text.

Some technologists, including Deepmind, believe that one day humans will develop “comprehensive” AI systems that will be able to work better or better than humans. Some call this artificial “normal” intelligence. Others say it’s like “believing in magic”. Many top researchers, such as Metana’s chief AI scientist Yan Lacan, question whether this is possible.

Gato is “normal” in the sense that he can do many different things at once. But it’s a world other than “normal” AI that can meaningfully adapt to new tasks that are different from the model that was trained, says Andres of MIT. “We are still a long way from being able to do that.”

Enlarging the model will also not address the issue that models do not have “lifelong learning”, meaning they can be taught things at once and they will understand all the implications and use them to inform the decisions they are going to make. Make, he says.

Emmanuel Cahmebay, part of the Black in AI organization, co-founded by AI / robotics researcher and Timnit Gabru, argues that the hype surrounding tools like GETO is detrimental to the normal development of AI. “There are a lot of interesting topics that have been left out, with less funding, which deserves more attention, but the big tech companies and most of the researchers at such tech companies are not interested in it,” he says.

Patrick J. Vilas Dhar, president of the McGovern Foundation, a charitable organization that funds AI projects “for good,” says tech companies should take a step back and take stock of why they’re building what they’re building.

“AGI communicates something in-depth to humans – the idea that we can be more than we are, by creating tools that propel us to greatness,” he says. “And that’s really nice, except it’s a way to distract us from the fact that we have real problems that we have today that we should try to solve using AI.”

Similar Posts

Leave a Reply

Your email address will not be published.