Forget ChatGPT, Elon Musk is making TruthGPT, an AI to 'understand the universe'

Elon Musk has officially announced his intention to clamber aboard the AI bandwagon with an alternative to ChatGPT which is apparently designed to be a "maximum truth-seeking AI" no less.

Musk spilled the beans on this artificial intelligence rival - which will also join Google Bard in the generative AI arms race - in an interview with Tucker Carlson (Fox News), as The Verge reports.

Musk told Carlson:

"I'm going to start something which I call TruthGPT or a maximum truth-seeking AI that tries to understand the nature of the universe.

And I think this might be the best path to safety in the sense that an AI that cares about understanding the universe is unlikely to annihilate humans because we are an interesting part of the universe."

As you may recall, this comes hot on the heels of Musk airing concerns about the direction AI tech is going in, and whether it's a dangerous road to continue down. That included the billionaire signing an open letter warning that as it stands, AI represents a threat to society and humanity.

So, it appears that we now know the direction Musk wants to take in order to correct the course of AI, and ensure it's not a bad thing (TM) that'll eventually culminate in the destruction of humanity (as is the case in most sci-fi novels, let's face it - if it's not annihilation, it's slavery).

This also follows recent news of Musk opening his own AI startup, X.AI Corp, last month, as well as buying an absolute boatload of GPUs and recruiting some AI experts, all of which makes sense as pieces of the puzzle coming together.

Up in the air

It seems TruthGPT is really happening, then, but as to how far the project has got, and what form it will ultimately take, that's all very up in the air (unlike a certain rocket). All we have right now is the vague assertions presented here, and the knowledge that some of the groundwork is in place - quite a lot of it, in fairness, and certainly enough to tell that Musk is serious about this ChatGPT rival.

Indeed, TruthGPT might have been in the works at least two months ago, based on a tweet in February where Musk mentioned it (see above), though this could have just been an off-the-cuff comment at the time for all we know (which was later grabbed for the name).

As The Verge points out, Musk also implied that OpenAI's incentive to turn a profit made for a difficult situation in the potential for that to interfere with the ethics side of its AI - and that TruthGPT would be a more transparent entity. We believe that's what they call 'fighting talk' in the trade...

As if the battle between Microsoft's ChatGPT-powered Bing AI and Google's Bard hasn't been hard fought enough thus far, we suspect this could really put the cat among the AI pigeons, as it were. Even if Musk is clearly behind his rivals in getting off the starting line, the resources seemingly being gathered already indicate that there might be a strong push coming to catch up.

Doubtless we'll be hearing a lot more about all this on Twitter soon enough.