Elon Musk is known for making some big claims but this time around, he’s getting a brutal response from a leading data scientist from Meta regarding AI.
The tech billionaire mentioned in a tweet regarding how AI would outsmart all of humanity by the year 2029. He was confident that this would happen but Meta disagreed.
Yann LeCun who is known to lead the AI industry at Meta mentioned to Musk how AI would not be able to compete with the human mind so soon and therefore refuted the claims made by Musk as false.
Musk and his love for AI is something unique, to say the least, and his enthusiasm can always be reflected through his words. This was certainly the case again but now that he’s getting outright criticized by Meta is definitely something he might not welcome.
Not only that, LeCun was quick to take a dig at Musk’s Tesla company and his desire to produce self-driving cars that he feels isn’t going as planned and might be overly ambitious in design.
Speaking on the X app, he says that if the idea was really brilliant, we would be seeing the world of AI teaching themselves a lesson on driving in just 20 hours, just like you’d see with a regular 17-year-old wanting a license.
Today, self-driving and reliable driving systems are not present, even if there are millions of databases featuring AI-labeled training materials, the chief Meta scientist added.
This is not the first time that we’ve seen Meta oppose Musk openly on a public forum. Jokes about how cats were more clever were even added and we’re not sure that Musk would love to hear that.
Meta added how we’ve been seeing so many tales about AI taking over the world and leaving humans behind or replacing them. But that’s not happening, at least anytime soon. There were even discussions in the past about how AI can be trained to give you the methodology to make chemical weapons but that again wasn’t true.
And on top of that, he mentioned how AI models have as much common sense as the brain of a housecat but they just were not as clever as them. Further discussions were about cats remembering and comprehending better and more about today’s physical world. At the same time, planning out complex scenarios and doing a certain degree of reasoning was similarly allowed. It’s much better than what we’re seeing with the biggest language models today, he explained.
There’s a massive ingredient that happens to be missing in the entire concept of machine learning and this might be what’s holding the large language models back, he added. The missing piece is on a grander scale and it just goes to show that they can never be as smart as animals or the human mind.
But there is still optimism about how these systems could someday give us all more useful data than what we’re seeing with the classic search engine approach. However, when and if that day is near, only time can tell but it’s not soon.
This is not what Musk agrees with. He has been super vocal about how he firmly believes that AI is this much powerful that it could be dangerous and stated how it might develop intelligence that overpowers the human mind out there today.
There’s a minor chance of it consuming humanity to such an extent that they become useless but he does feel that likelihood is not zero. He does feel AI advancements can however put a constraint on the potential of the growth of the human mind.
For years, the Tesla owner has put his money on AI and machine learning and that is evident whenever he speaks about the technology in public during his speeches. Do you remember his over-ambitious plans of rolling out robotaxis as early as the year 2020? Thankfully, that goal was pushed back.
Musk also happens to be a bigger cofounder and ex-board member of the OpenAI firm which LeCun has been very vocal about since day one.
Read next: Viral Social Media Content Sparks Outrage and Threat Perception, Research Finds
