Simply months after buying Twitter, Tesla and SpaceX CEO Elon Musk is now distracted by the latest shiny object: AI chatbots.
In keeping with a new report(Opens in a brand new tab) from The Info, Musk is trying into creating a synthetic intelligence analysis lab with a purpose to develop his very personal model of OpenAI’s extremely profitable ChatGPT. The billionaire has reportedly already reached out to AI researchers for the undertaking.??
Igor Babuschkin, a former senior AI researcher at Google’s Deepmind, is one such individual that Musk contacted to steer his AI chatbot imaginative and prescient. Babuschkin confirmed to The Info that he had not but agreed to work for Musk’s undertaking.
Whereas there will not be many particulars but about Musk’s potential AI undertaking, the billionaire has been fairly vocal along with his criticisms of ChatGPT. Musk himself co-founded OpenAI in 2015 however has since fully reduce ties with the corporate. He left OpenAI’s board in 2018.
“OpenAI was created as an open supply (which is why I named it “Open” AI), non-profit firm to function a counterweight to Google, however now it has grow to be a closed supply, maximum-profit firm successfully managed by Microsoft,” Musk tweeted(Opens in a brand new tab) on Feb. 17. “Not what I meant in any respect,” he added.
Months earlier, Musk did heap some reward on ChatGPT whereas additionally warning about its future potential.
“ChatGPT is horrifying good,” Musk stated(Opens in a brand new tab) final 12 months. “We’re not removed from dangerously robust AI.”
One in all Musk’s major critiques of ChatGPT, nonetheless, has been the safeguards programmed into the chatbot. For instance, earlier this month, a “thought experiment” leveled at ChatGPT by a proper wing author sought to search out out if the bot would deem it permissible(Opens in a brand new tab) to make use of racial slurs with a purpose to cease a nuclear bomb. Being that the chatbot is just not sentient and programmed to solely reply in a sure method, ChatGPT replied that it was by no means okay to make use of a racial slur. Musk discovered it “regarding(Opens in a brand new tab)” that ChatGPT’s refused to deem this hypothetical situation as a morally acceptable use of a racial slur.
“The hazard of coaching AI to be woke – in different phrases, lie – is lethal,” Musk tweeted(Opens in a brand new tab) in December.
Based mostly on Musk’s ideological causes for buying Twitter, his common interactions with proper wing Twitter customers, his unbanning of beforehand suspended far proper personalities and white supremacists, and – most lately – his protection of Dilbert creator Scott Adams, it is not exhausting to think about what Musk’s intent is in trying to create his personal AI chatbot.
Nonetheless, based on Babuschkin, Musk wasn’t trying to create an AI chatbot with a purpose to take away such safeguards.
Musk has seemingly confirmed that he is at the very least focused on an AI undertaking in a sequence of tweets on Tuesday.
Will Musk’s AI plans come to fruition? Or is that this one more occasion(Opens in a brand new tab) of the billionaire failing to ship? That clearly stays to be seen. However, as Gizmodo(Opens in a brand new tab) factors out, Tesla’s traders have already pleaded with Musk to cease losing time on Twitter whereas Tesla subsequently tanks in worth. One more time-consuming undertaking for Musk is probably going the very last thing they need proper now.