A lot has happened in the last year regarding Twitter, Elon Musk and Artificial Intelligence. But if something has remained constant and certain, it is that the head of Twitter, Tesla or SpaceX does not seem to take being in the background very well.
In April of last year, Elon Musk began his assault on buying Twitter (which we now know he regrets) and only a few months later the world began to explore and, in a way, be amazed by developments like OpenAI’s ChatGPT. and its possibilities.
Perhaps for this reason (although without neglecting his genuine interest in the subject) in recent days we have learned that the tycoon’s next two initiatives are going to be closely linked to AI. We talk about X.AIwhich appears to be the parent corporation of several of his companies and interests, and TruthGPTits own and heralded competitor to ChatGPT with a rather thinly disguised name.
What implications can this have in the medium term? Is it one of the more Musk announcements that can come to nothing or does it seem serious? What would that chat be like, stubborn in supposedly always telling the truth? How will this affect Twitter, Tesla or the rest of their companies?
We have compiled everything that is known so far about these two new initiatives:
What is X.AI?
Musk incorporated a company called X.AI on March 9., according to Nevada business records. He is the only director of the company, his secretary is listed as Jared Birchall, the former Morgan Stanley banker who manages part of Musk’s wealth.
Birchall is also the CEO of Musk’s chip implant company, Neuralink, and is on the board of Musk’s The Boring Company. He is also the CEO of the Musk family office, Excession. Come on, your consigliere particular.
After this founding, Musk recently changed Twitter’s name to X Corp on company records, as part of its plans to create an “everything app” under the “X” brand.
Although little is known about the company so far, Musk has acknowledged that it is buying high-powered NVIDIA GPUs to pursue generative AI, the field on which chatbots like OpenAI’s ChatGPT are based.
But what would X be and where does this concept come from? X was already the name that Musk gave to one of his first big companies, a fintech that would end up being the seed of PayPal. After owning it, a few years ago he bought back the x.com domain for an undisclosed sum that was sure to be significant. Now, it is known that he has acquired the X.ai domain, although again without knowing the amount.
X is Musk’s vision of an “app for everything”a single piece of software that can be used for all the purposes of everyday life and work, from social media to payments and, apparently, artificial intelligence.
What is TruthGPT?
Shortly after learning this, last Thursday, Elon Musk assured that he is working on “TruthGPT”, an alternative to ChatGPT that acts as a “Ultimate truth-seeking AI”. The billionaire laid out his vision for this rival AI during an interview on Fox News, stating that an alternative approach to creating AI was necessary to prevent the destruction of humanity.
I’m going to run something I call TruthGPT or an AI that seeks the ultimate truth and 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 kill humans because we’re an interesting part of the universe.
Elon Musk
From the network in which he made the announcement, and from some of the latest opinions that Musk has been leaving, it seems that his possible AI would be an attempt, from his point of view, to these chatbots will avoid biases marked by the so-called movement woke or excessive political correctness.
Asked about the supposed risks of unrestricted AI —after it is known that OpenAI or Microsoft have limited the possibility that their technologies answer certain questions after cases such as DAN—, Musk alluded to the fact that the lack of an AI’s desire to destroy all of humanity with the way humans strive to protect chimpanzees. “We recognize that humanity might decide to hunt down and kill all chimpanzees,” Musk said, but “we’re actually glad they exist, and we aspire to protect their habitats.”
In the interview with Fox News, Musk, normally brash and short-sighted with his promises, seemed to be on guard when discussing his bets on his pursuit of AI. Explaining why he was entering a field in which Google and Microsoft already have a head start, he replied that:
I think I’ll create a third option, although I’ll start very late of course. Definitely a late start. But I will try to create a third option and that third option I hope will do more good than harm.
Elon Musk
Also, we know you are hiring engineers from the best AI labs, like Igor Babuschkin, former DeepMind employee, and half a dozen engineers. The Information already reported the first talks of Babuschkin with Musk.
It’s not the first time Musk has pondered creating a “TruthGPT.” In February he tweeted that “what we need is TruthGPT,” while calling attention to the risks of large-scale AI models, such as those created by Open AI.
It should be remembered that Musk, along with other AI researchers, signed an open letter in March urging companies to stop “giant AI experiments” that their creators cannot “reliably understand, predict, or control.”
That letter was published just days after opening his new company in Nevada.
What position had Elon Musk maintained so far with AI?
Musk was an early donor of OpenAI, initially conceived as a non-profit organization. He, too, served on his board of directors before leaving in 2018. In his interview with Fox, he explained how the AI project he helped start got out of hand.
“I really put a lot of effort into creating this organization to serve as a counterweight to Google,” Musk said. “And then I sort of took my eye off of it, I guess, now they’re closed source and obviously for-profit.”
Musk has been critical of OpenAI and the direction it has taken under its CEO, Sam Altman.and has decried the limitations it has placed on AI, which Musk sees as a violation of the truth.
“My assessment of why AI is overlooked by very smart people is that very smart people don’t believe that a computer can ever be as smart as they are,” Musk said in an interview with The New York Times in the summer of 2020.
In 2017, Musk clashed with Facebook CEO Mark Zuckerberg when the latter called Musk an AI denier and doomsday prophet, lor that led Musk to deny Zuckerberg’s expertise in AI.
But beyond his criticisms, Musk’s companies are also experimenting with AI. In particular Tesla with its Full Self-Driving autonomous driving software. Musk promised in 2019 that Tesla vehicles would one day “wake up” and become autonomous, a promise that has yet to come true.
Will Twitter be part of X.AI from now on? And Tesla or SpaceX? How does it affect Musk’s plans to create a SuperApp?
Yes, in fact, as we have seen before, it already is. This is not the case with Tesla (unfeasible as it is a listed company) or SpaceX. Musk already covers five companies in which he actively participates, something criticized by the shareholders of his most profitable companies.
Musk is caught between a cycle of putting out daily fires on Twitter, where last week he admitted to sleeping on a couch on the seventh floor, and balancing responsibilities at Tesla and SpaceX that literally amount to projects focused on getting to the Moon and changing the world. the way we transport ourselves.
But after unveiling his new plans, the original idea that comes to mind is to think of a possible Twitter evolved to X as something very similar to the Chinese super application WeChat.
But Musk’s aspiration to make Twitter a SuperApp also means that it is profitable. Let’s make it clear beforehand that Twitter has always had a relationship far from profitable. And Musk, if something seems obvious, is that he can do it better or worse, but he doesn’t want to lose money. Hence his supposed intentions to improve functionalities, his somewhat fraudulent commitment to Twitter Blue and many other movements.
Twitter has been the great social network that has had the most difficulty being profitable. He did it in 2018, he did it again in 2019, and in 2020, due to the drop in advertising due to the pandemic, he returned to the red. All while its users seemed to remain flat.
This SuperApp, if it ends up happening, could start from Twitter as the center to deploy from it an ecosystem of services that could be more profitable.
Can Musk use what we post on Twitter to feed his AI?
The truth is that yes, but this is not a novelty of the Musk era. In fact, the content generated on Twitter was already used and sold to create big data and market studies.
The purchase of Twitter could prove useful in helping to train its new AI, whose systems rely on reams of information on the Internet to build models that mimic human speech and writing.
Do we agree with this? We will have to see if some protection formula is offered that does not involve leaving Twitter if we do not want our tweets to feed this Musk AI but, for now, the truth is that systems like ChatGPT had already been drinking from Twitter until Elon himself turned off the tap.