Sunday, October 1, 2023

“AIs still cannot reason or plan”

What will happen after ChatGPT and Midjourney? Behind the rise of generative artificial intelligence (AI) tools, capable of creating texts and images, what are the giants of the sector working on? at the party world, On Sunday, September 17, Naila Murray spoke about the research avenues of Meta, the parent company of Facebook and Instagram, where she runs artificial intelligence laboratories in Paris, London, Zurich and Tel Aviv.

Read also: Article reserved for our AI subscribers: Meta puts pressure on OpenAI and Google with an open source model

The main language processing models, on which many tools are based, including ChatGPT, “have statistical logic” and “I still can’t reason and plan,” pointed to Mme Murray. One way to explain that this software is limited to creating realistic texts, similar to the billions of pages they have ingested. Therefore, in its fundamental research, Meta works on AIs that would be endowed with these capabilities and that could forge a “representation of the world.”

Behind this abstract expression lies the idea of ​​allowing software to learn things, without human supervision, by watching videos or even observing reality, through video sensors. “This is inspired by the ability of humans or animals to understand how the world works, which allows them to make predictions: if they see a person walking, they expect that person to put their foot on the ground because they understand what it is like.” walk. . If they see me throw a ball in the air, they know it’s going to come down, etc. “, explained Murray.

Read Also:  Prime Minister meets with representatives of the Vietnamese community in the United States

Software that imitates human “common sense”

With this in mind, Yann LeCun, director of artificial intelligence research at Meta, launched I-Jepa in June, an attempt to create a new AI model capable of learning, by ingesting images, abstract representations of things (the paws of a dog , a dog’s forehead, etc.), without making the effort to analyze each pixel, as current image generation software does. The goal is to imitate the ” common sense “ human. And ultimately create machines that “Learn much faster, plan how to perform complex tasks and adapt to unexpected situations.” explains the company. LeCun, one of the French researchers known as one of the fathers of modern AI, has sometimes given the example of software capable, if asked to organize a Paris-New York trip, of indicating the exit of the building, ordering a taxi, to guide us to the airport, buy a plane ticket, etc.

Read Also:  More than 10,000 new computers for students in Nova Scotia

You have 52.19% of this article left to read. The rest is reserved for subscribers.

Times of National
Times of National
Times of National To give more information about the latest happenings, news related to happenings in the country and abroad a casual understanding of the latest technology products and gadgets, celebrity news and gossip, latest movie news, sports, and cricket scores all you need Always ready to fulfill whatever else is becoming a part of our life nowadays.
Latest news
Related news


Please enter your comment!
Please enter your name here