Twitter CEO Elon Musk is reportedly exploring options for a generative AI project for Twitter, with plans to develop his own AI tools to compete with OpenAI. As part of this initiative, Musk has purchased thousands of GPUs to fuel the project.
According to Insider, Musk is exploring the use of large language models (LLMs) for his AI project at Twitter. Essentially, LLMs are AI systems that are trained on vast amounts of data so that they can produce new, seemingly sophisticated content and text independently.
Twitter has an abundance of data that could potentially be used to train LLMs. Previously, OpenAI had used Twitter data to train its GPT models, but Twitter cut off its access last year.
Now, Musk and his team are developing their own competing project to harness the vast amount of data that Twitter has to offer.
Musk’s push into AI is motivated by his belief that AI tools can be agenda-shifting models. This is exemplified by his recent move to recruit AI developers to create his own ‘anti-woke’ AI model.
He is also seeking to build his own AI tools to compete with OpenAI. One reason behind this is his frustration with OpenAI becoming a for-profit business after receiving $100 million from Musk during its early stages as a non-profit.
Musk’s AI push is likely to be motivated by his broader goal of promoting free speech and providing a platform for all viewpoints on Twitter. It’s not yet clear how AI will fit into Twitter, but Musk’s exploration of LLMs suggests that he sees potential for AI to revolutionize the platform.
However, the idea of an AI model built entirely on tweets raises concerns about the potential for amplifying the negativity and toxicity that are already pervasive on the platform.