Meta AI chief is ‘not impressed’ with ChatGPT, here’s why – Times of India

ChatGPT, the AI chatbot from OpenAI, has enjoyed a lot of fanfare and attention since rolling out. A day doesn’t pass without the mention of ChatGPT in the news. Experts have suggested that ChatGPT could be the next big thing in tech. However, in the eyes of Meta chief AI scientist Yann Le Cun, it is “nothing revolutionary” and just “well put together.”
As reported by ZDNET, Le Cun shared his thoughts about ChatGPT during a Zoom call organized by Collective[i] Forecast last week. The AI chief, who has a Turing award among his accolades, talked about the commonality of ChatGPT-like AI in the industry and why no tech giant has released something similar till now.

ChatGPT not alone in using such AI technologies, built upon those developed by other firms
According to Le Cun, ChatGPT isn’t the pioneer it may otherwise seem to be because of popular opinion. He said, “In terms of underlying techniques, ChatGPT is not particularly innovative … It’s nothing revolutionary, although that’s the way it’s perceived in the public.”
Meta’s AI head noted that the technologies leveraged by ChatGPT’s owner, OpenAI, are common in this domain. As per Le Cun, many companies employ AI techniques that are similar to those used by OpenAI. He revealed, “OpenAI is not particularly an advance compared to the other labs, at all … It’s not only just Google and Meta but there are half a dozen startups that basically have very similar technology to it.”

The AI scientist also mentioned that GPT-3, the underlying base of ChatGPT, is a consequence of multiple technologies, built by many different companies, that came into existence over decades. He continued, “You have to realize, ChatGPT uses Transformer architectures that are pre-trained in this self-supervised manner … Transformers is a Google invention.” He added, “So there’s a whole history of this, and this did not come out of a vacuum.”
This is why Meta or Google have not released a ChatGPT competitor yet
Talking about tech giants and their lack of public releases in this area, Le Cun said, “Google and Meta both have a lot to lose by putting out systems that make stuff up.” Despite this risk, he confirmed the Zuckerberg-led company will implement AI for a variety of purposes in the future. He shared that the company’s plan goes beyond “text generation” and into “creation aids,” such as “generative art.”
Also watch