ChatGPT is fashionable, very fashionable. Articles, videos, analysis… in recent weeks thousands and thousands of contents have been published dedicated to the chatbot created by OpenAI and which has amazed half the world. We, you will remember, decided to put it to the test to find its weak points, and we have also fixed our gaze on its darker face, while we told you about Microsoft’s interest in this service and how there are those who are trying to profit from apps from access to it.
Personally, and I have already commented on it on more than one occasion, ChatGPT seems to me to be a truly innovative proposal, since until now we have never seen an artificial intelligence capable of offering that level of interaction, exclusively based on text, with human beings. Many of his answers are correct and, what is just as important, he knows how to adapt their format and level to the interlocutor, thus becoming a universal interlocutor.
However, I read on ZDNet that Yann LeCun, an institution in the world of artificial intelligence and head of AI at Meta, states that ChatGPT is not “particularly innovative» nor “nothing revolutionary”«. This is what we can read in that article:
« “In terms of the underlying techniques, ChatGPT is not particularly innovative,” Yann LeCun, Meta’s chief AI scientist, told a small meeting of press and executives on Zoom last week.
“It’s nothing revolutionary, even though that’s how it’s perceived by the public,” LeCun said. ‘It’s just, you know, it’s well put together, it’s very well done’ «.
I think it’s important to remember that LeCun was awarded in 2019 with the Turing Award together with Geoffrey Hinton and Yoshua Bengio, so we are talking about a most relevant personality in this field, with a level of knowledge in the matter far superior to that of the vast majority of mortals, and therefore an opinion that is important to listen And from that perspective, I don’t consider myself the one to discuss his argument about the technology on which ChatGPT is based, because I am totally convinced that he is right.
On the technical side.
Now, only the technology used determines whether or not something is innovative? All development is based on previously developed technologiesso if we start from that premise, nothing is really innovative, since everything is part of a chain of pre-existing technologies, right?
ChatGPT is innovative because it makes a previously unseen use of these technologies that, although already veteran, had not been used to create a general-purpose chatbot with universal access, capable of adapting its responses to its interlocutor. And even if the infrastructure on which it runs is based on steam engines, it would still be an innovative proposition.
Google’s need to provide a quick response, Microsoft’s interest in integrating it into its products and services, the interest aroused by future payment accounts, the level of saturation of its servers… there are many elements that indicate that yes, that ChatGPT really is an innovative proposal, and a lot. That the technology on which it is based already existed decades ago? Well, that it was possible to do it but no one has done it until now says more than who points a finger.
The ChatGPT input, is it or is it not so innovative? It was first published on MuyComputer.
