Elon Musk says xAI's Grok will become open source and calls OpenAI 'a lie'

The move will be made this week as the market competition heats up

Mr Musk has been in a bitter public spat with OpenAI since February 29 when he filed a lawsuit against the company. Reuters
Powered by automated translation

Billionaire businessman Elon Musk has said his artificial intelligence company xAI will make its Grok chatbot open source, doubling down on his offensive against ChatGPT maker OpenAI.

The move will be made this week, Mr Musk said in a post on X, formerly Twitter.

Open-sourcing, by definition, means it should be free, including its redistribution. But there are exceptions like, for example, if the software is too restrictive and cannot be utilised for personal use, or if there are trademark considerations involved.

The governing Open Source Definition, published by the Open Source Initiative, focuses on the software's source licence rather than on what users are free to do.

Richard Stallman, the founder of GNU Project, highlighted this conflicting duality in his blog post titled Why Open Source Misses the Point of Free Software.

“That causes confusion in paradoxical situations where the source code is open source [and free] but the executable itself is non-free,” he said.

Grok is currently only available to subscribers of X's Premium+ tier, technically meaning users still have to pay to access Grok's open-source code, and it is unclear what they can or how far they can go with their access to Grok.

Mr Musk has been in a bitter public spat with OpenAI since February 29 when he filed a lawsuit against the company.

The case alleges that OpenAI has deviated from its original mission of providing AI access to all and has instead pursued profits.

OpenAI has denied the allegations, stating that Mr Musk regrets not being part of the company and its success today, as it has morphed into the leader of the generative AI revolution.

On Monday, Mr Musk continued the row on X, calling OpenAI a “lie”.

OpenAI has yet to respond to xAI's plans.

What is Grok?

Grok is xAI's generative AI chatbot, based on a large language model, the underlying algorithm that uses deep learning and analyses significant amounts of data to generate content.

Its website simply describes Grok as a “conversational AI for understanding the universe” – aligned with Mr Musk's mission statement when he announced in July last year that xAI was meant “to understand reality” and “the true nature of the universe”.

Grok is also “designed to have a little humour in its responses … and loves sarcasm. I have no idea who could have guided it this way”, he said.

“It will also answer spicy questions that are rejected by most other AI systems,” according to xAI's website.

How does Grok differ from its rivals?

While all generative AI chatbots rely on LLMs, Grok has one distinct edge, according to Mr Musk.

It has real-time access to information via the X platform, which is a “massive advantage over other models”, Mr Musk said.

Grok's main LLM is Grok-1, an evolution of the original Grok-0, which was trained on 33 billion parameters. Grok-1 has gone through several iterations over the past four months and is “significantly more powerful”, xAI said.

According to xAI, Grok-1 still lags behind OpenAI's GPT-4 and Anthropic's Claude 2 in benchmarking statistics.

Will open-sourcing Grok be an advantage?

Mr Musk is no stranger to championing open-sourcing. During his acquisition of Twitter in 2022, one of his pitches was that he would make the platform's code open-source.

He has not given any reasons why he decided to open-source Grok, but the move does come with its advantages.

Open-sourcing code is meant to promote transparency, allowing the public to inspect and scrutinise proprietary software and pitch their ideas to developers on how to change codes or even use the algorithm.

This would mean Grok will have a bigger pool of developers – or even ordinary people – who can flag their suggestions, concerns and insights, which in turn, would help xAI improve its platform.

US Senate holds hearing on regulating AI

OpenAI CEO Sam Altman, from right, NYU Professor Emeritus Gary Marcus and IBM Chief Privacy and Trust Officer Christina Montgomery are sworn in before testifying at a Senate Judiciary Subcommittee on Privacy, Technology and the Law hearing on artificial intelligence, Tuesday, May 16, 2023, on Capitol Hill in Washington.  (AP Photo / Patrick Semansky)

That also falls in line with Grok's description, stating that “human feedback is essential”, despite acknowledging that providing consistent and accurate feedback “can be challenging”.

“AI can assist with scaleable oversight by looking up references from different sources, verifying intermediate steps with external tools and seeking human feedback when necessary,” xAI said.

Is Grok already available?

Grok is only available to subscribers of X's Premium+ tier, which starts at $16 per month, or $168 a year. It comes with revenue sharing, gives users a large reply boost, grants access to other creator tools and removes advertising from feeds.

X has not revealed its Premium+ subscriber numbers. However, the latest data from Statista shows that there were about 640,000 subscribers as of April last year. A more recent study from data analyst Travis Brown said there were about 892,000 as of September.

What's the latest in the Musk-OpenAI feud?

After OpenAI told its employees in a memo that Mr Musk, a co-founder and early backer of OpenAI, regrets not being part of the company – just days after filing the lawsuit – OpenAI last week published his emails, shedding light into the early days of their partnership.

OpenAI said that Mr Musk did not believe that the company would be successful and that it needed to raise a significant amount of funds to succeed in the sector.

“We're sad that it's come to this with someone whom we’ve deeply admired … and then sued us when we started making meaningful progress towards OpenAI’s mission without him,” the company said.

On Monday, OpenAI doubled down on its retaliation, saying in a California court filing that there was no founding agreement, “or any agreement at all with Musk”, formally denying the billionaire's claims it initially made in its staff memo.

OpenAI said Mr Musk's allegations “rest on convoluted – often incoherent – factual premises”, and the “relief Musk seeks is as extraordinary as his claims are contrived”.

“Musk requests an order compelling OpenAI to reorganise and distribute its technology in accordance with the terms of his fictitious contract,” the filing said.

Updated: March 12, 2024, 9:24 AM