Billionaire Elon Musk recently revealed his plans of establishing a new artificial intelligence (AI) firm with the aim of challenging Microsoft-backed OpenAI. While OpenAI has been making waves in the fast-evolving AI domain with the impressive ChatGPT, tech giant Google has also been trying to catch up with Microsoft’s AI dominance by coming up with its own generative AI chatbot. While admitting that he’s late to the race, Musk said he wants to create a ‘third option’ that can take on ChatGPT and Google’s Bard.
Although ChatGPT is ahead of Bard in terms of accuracy at the moment, many experts have raised concerns about ChatGPT’s functioning, some of them from an ethical standpoint. Musk has also been very critical of rapid developments within AI research in general, and ChatGPT’s evolution specifically. The Tesla and Twitter boss was also one of the key signatories of an open letter that called for a complete pause on advanced AI research, warning of “profound risks to society and humanity”. The letter, signed by more than 1,000 leaders in the tech domain including Apple co-founder Steve Wozniak, requested that such a pause should end only when the industry is confident that risks associated with advanced AI systems will be manageable.
Further, Elon Musk took his opposition to Microsoft’s hold over OpenAI to a new level when he threatened to sue Microsoft earlier this month. His vague accusation of “illegal training using Twitter data” seems to be about OpenAI’s use of Twitter data to train the language models behind ChatGPT. It should be noted that this was not the first time that Elon Musk publicly expressed his concerns around OpenAI’s functioning or Micrsoft’s alleged control over it. The reason behind his distress can be understood better if one looks at his close association with OpenAI in its initial years.
Musk’s OpenAI Mishap
OpenAI is an AI research laboratory that was completely non-profit in nature when it began operations in 2015. Back then, it had the backing of tech billionaires like Elon Musk and Reid Hoffman, and it was designed as entirely open source in structure, and hence the name. However, after having served as a member on the firm’s board since its inception and having donated big sums to the non-profit research lab, Musk had a falling out with OpenAI in 2018. He left all positions in the firm that year and continued to observe its actions as a ‘neutral observer’, as per his own words.
Although there are several speculations around why Musk quit OpenAI, the official version was that since Tesla was also doing its own AI research, Musk did not want any conflict of interest. According to various reports, Musk wanted to completely take over OpenAI and run it on his own to catch up with Google’s AI research. However, this was not to be.
Since then, Musk reportedly went back on the payments he had earlier promised to OpenAI. The billionaire also expressed concern at OpenAI shedding its non-profit nature and pivoting towards a revenue generating mechanism. As Microsoft started investing more and more in the AI research lab, the tech major’s grasp over the AI domain also tightened. When OpenAI gave exclusive license to Microsoft for its GPT-3 language model, Musk commented that OpenAI was essentially captured by Microsoft.
In a more recent episode, Musk said, “OpenAI...has become a closed source, maximum-profit company effectively controlled by Microsoft.” Notwithstanding what Musk or others said about its not-so-open nature, OpenAI became more prominent since Musk’s departure. Its interactive chatbot ChatGPT and image generator DALL-E has changed the way people view generative AI.
From what Musk has been repeatedly saying, OpenAI would have been functioning in a very different manner if he still held any influence over the firm. The current trajectory of ChatGPT’s evolution will take the chatbot into a position that Musk termed as ‘dangerously strong’. This view adds up with his take on AI regulation as well. Other than the recent letter which called for a moratorium on advanced AI research, Musk has been advocating for strict regulation of AI for a very long time now.
Musk is in no position to exert any influence over OpenAI right now and Microsoft seems to be in the driver’s seat as far as world’s most advanced AI research is concerned. While his new venture ‘TruthGPT’ might want to challenge ChatGPT’s dominance, not much is known on how the Tesla boss plans to achieve that. In a domain where even Google has fallen behind as of now, it will not be easy for Musk’s new initiative to make a substantial mark.
As it stands, Musk’s hostility towards Microsoft’s control over OpenAI and the research lab’s functioning will likely continue as it is. Even though the tech billionaire has ambitions around TruthGPT, he has other pressing concerns to address such as the downgrading of Tesla’s stock, falling advertising revenue at Twitter and the recent SpaceX crash.