In terms of underlying methods, OpenAI’s ChatGPT chatbot is not “particularly innovative” and “nothing revolutionary” about it. This was stated by Vice President of Meta and leading AI specialist of the company Jan LeKun at a Zoom meeting with the press, writes ZDNet.
“It’s just well put together, beautifully done,” the researcher said.
According to LeCun, the idea that OpenAI is “unique in its kind of activity” is wrong. The AI lab is “not very advanced” compared to other companies, he added.
“It’s not just Google or Meta, but half a dozen startups that have similar systems,” the scientist noted.
LeCun pointed out that ChatGPT and the language model it is based on, GPT-3, includes several third-party technologies.
The chatbot is based on pre-trained self-controlled transformers, which were first described by Google Brain engineers in 2017. These neural networks formed the basis of many language systems.
“And self-supervised learning is a topic that I have been talking about for a long time, even before there was OpenAI,” LeCun said.
He added that the first big language model appeared about 20 years ago. It was created by the head of the Montreal AI Research Institute MILA, Yoshua Bengio.
Also, OpenAI developers used the reinforcement learning method, which was first used by researchers from the Google DeepMind division.
“Him [ChatGPT] there’s a whole story, he didn’t come out of a vacuum,” LeCun said.
According to a top Meta manager, the OpenAI chatbot is not so much a scientific breakthrough as “a worthy engineering approach.”
“And I’m not going to criticize them for that,” he said.
Asked by a New York Times reporter if the public would ever identify Meta’s breakthrough AI team FAIR as OpenAI, LeCun answered yes.
“Not only about text generation, but also about creativity,” the scientist added.
The researcher was also asked about the reasons for the lack of ChatGPT-like technologies from Google and Meta.
“They are [техгиганты] have a lot to lose by releasing algorithms that can invent something,” he replied.
Recall that in October 2022, Jan LeKun proposed to endow AI systems with “common sense”.
In February, OpenAI co-founder Ilya Sutskever admitted that artificial intelligence has consciousness. Yang LeKun categorically criticized his statement.
Subscribe to Cryplogger news in Telegram: Cryplogger AI – all the news from the world of AI!
Found a mistake in the text? Select it and press CTRL+ENTER
In terms of underlying methods, OpenAI’s ChatGPT chatbot is not “particularly innovative” and “nothing revolutionary” about it. This was stated by Vice President of Meta and leading AI specialist of the company Jan LeKun at a Zoom meeting with the press, writes ZDNet.
“It’s just well put together, beautifully done,” the researcher said.
According to LeCun, the idea that OpenAI is “unique in its kind of activity” is wrong. The AI lab is “not very advanced” compared to other companies, he added.
“It’s not just Google or Meta, but half a dozen startups that have similar systems,” the scientist noted.
LeCun pointed out that ChatGPT and the language model it is based on, GPT-3, includes several third-party technologies.
The chatbot is based on pre-trained self-controlled transformers, which were first described by Google Brain engineers in 2017. These neural networks formed the basis of many language systems.
“And self-supervised learning is a topic that I have been talking about for a long time, even before there was OpenAI,” LeCun said.
He added that the first big language model appeared about 20 years ago. It was created by the head of the Montreal AI Research Institute MILA, Yoshua Bengio.
Also, OpenAI developers used the reinforcement learning method, which was first used by researchers from the Google DeepMind division.
“Him [ChatGPT] there’s a whole story, he didn’t come out of a vacuum,” LeCun said.
According to a top Meta manager, the OpenAI chatbot is not so much a scientific breakthrough as “a worthy engineering approach.”
“And I’m not going to criticize them for that,” he said.
Asked by a New York Times reporter if the public would ever identify Meta’s breakthrough AI team FAIR as OpenAI, LeCun answered yes.
“Not only about text generation, but also about creativity,” the scientist added.
The researcher was also asked about the reasons for the lack of ChatGPT-like technologies from Google and Meta.
“They are [техгиганты] have a lot to lose by releasing algorithms that can invent something,” he replied.
Recall that in October 2022, Jan LeKun proposed to endow AI systems with “common sense”.
In February, OpenAI co-founder Ilya Sutskever admitted that artificial intelligence has consciousness. Yang LeKun categorically criticized his statement.
Subscribe to Cryplogger news in Telegram: Cryplogger AI – all the news from the world of AI!
Found a mistake in the text? Select it and press CTRL+ENTER
