Elon Musk plans to challenge ChatGPT and Google with 'truth-seeking' AI platform

Billionaire owner of Twitter says his TruthGPT platform 'might be the best path to safety' to curb dangers of AI

Elon Musk made the announcement during his interview with Fox News journalist Tucker Carlson. AP

Twitter chief executive Elon Musk plans to develop a new generative artificial intelligence platform to challenge Microsoft-backed OpenAI’s ChatGPT and Google’s Bard.

The platform, which the billionaire calls TruthGPT, would become a "third option" to what he calls "the two heavyweights" currently dominating the market that has become the new darling for investors, according to excerpts of Mr Musk's interview with Fox News Channel's Tucker Carlson.

"I'm going to start ... a maximum truth-seeking AI that tries to understand the nature of the universe," Mr Musk told Mr Carlson in the interview that aired on Monday.

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

He declined to provide specific details on his plans for developing TruthGPT.

Mr Musk's announcement confirms earlier reports by The Information in March and the Financial Times last week on his plans to throw his hat into the generative AI race.

San Francisco-based technology news site The Information said Mr Musk was assembling a team, which includes a former engineer at a unit of Google parent Alphabet, to develop a rival to ChatGPT.

It said he was in discussions with Igor Babuschkin, who recently left Google's DeepMind AI, to lead a group of artificial intelligence researchers in the endeavour. It remains unclear whether any of the other persons who have been approached have signed up so far.

The British daily FT, meanwhile, reported that Mr Musk is building a team of AI researchers and engineers, and also in talks with various investors in his other ventures, such as SpaceX and Tesla, about pumping capital into his new AI start-up.

He has also ordered “thousands of high-powered GPU [graphic processing unit]” chips from technology company Nvidia, the report said. GPUs are sophisticated chips capable of running many tasks simultaneously and are necessary in building generative AI’s large language models.

Last month, he reportedly merged registered a company named X.AI Corp in Nevada, according to a US state filing.

Generative AI can produce various kinds of data, including audio, code, images, text, simulations, 3D objects and videos. While it takes cues from existing data, it is also capable of generating new and unexpected outputs, according to GenerativeAI.net.

The global generative AI market is expected to reach $188.62 billion by 2032, growing at an annual rate of more than 36 per cent from $8.65 billion last year, according to Brainy Insights.

It could also drive a 7 per cent — or almost $7 trillion — increase in the global economy and lift productivity growth by 1.5 percentage points over a 10-year period, Goldman Sachs estimated.

Investors have poured in more than $4.2 billion through 215 deals into generative AI start-ups in 2021 and 2022 after interest spiked in 2019, data from CB Insights showed. Around $586 million in 20 deals during the same period has gone to generative interfaces, it said.

Mr Musk, who co-founded the California-based start-up OpenAI in 2015, left its board in 2018 following disagreements with the company's direction. However, in 2019, he tweeted that the need to focus on Tesla and SpaceX was the reason he left OpenAI.

He has also been critical of the company in recent months, arguing that OpenAI is placing several safety nets to prevent ChatGPT from offering results that might be divisive or offend its users. That implies that his planned TruthGPT chatbot might have less restrictions.

ChatGPT was launched in November by OpenAI — the world's most valuable generative AI start-up — and quickly gained popularity as it can generate written content with a simple request within seconds.

AI is more dangerous than, say, mismanaged aircraft design or production maintenance or bad car production

Elon Musk,
chief executive of Twitter and Tesla

In January, technology company Microsoft announced the third phase of its long-term partnership with OpenAI through a new multiyear, multibillion-dollar investment, worth a reported $10 billion.

Earlier this month, Google, meanwhile, unveiled its own conversational AI service, Bard. However, days later, Bard made an error in a promotional video during a company event in Paris, causing $100 billion to be wiped off Google's market value.

Mr Musk had criticisms for both companies during the interview with Mr Carlson, saying that OpenAI is "training the AI to lie" and that it has become a "closed source", "for-profit" platform "closely allied with Microsoft".

He also accused Google co-founder Larry Page of not taking AI safety seriously and that their disagreements were the reason OpenAI now exists.

"The reason OpenAI exists at all is that Larry Page and I used to be close friends and I would stay at his house in Palo Alto, and I would talk to him late into the night about AI safety," Mr Musk said.

"And at least my perception was that Larry was not taking AI safety seriously enough."

It is unclear how much Mr Musk, the world's second-wealthiest person, is willing to invest in his chatbot venture. He paid $44 billion to acquire Twitter last year.

"Conservational AI is currently in its early stages of monetisation and costs remain high as it is expensive to run," Swiss bank UBS said in a research note.

Mr Musk also remained cautious about the risks of AI in his interview with Mr Carlson, comparing it to other industries that now widely utilise the technology.

"AI is more dangerous than, say, mismanaged aircraft design or production maintenance or bad car production," he said.