Elon Musk propels Twitter’s AI revolution with a massive GPU arsenal
Elon Musk pushes ahead with a Twitter‑based AI initiative
Despite recently calling for a pause in AI training across the industry, the Tesla billionaire appears to be moving forward with an artificial‑intelligence project within Twitter. The platform, which Musk acquired almost six months ago, has already seen dramatic changes, yet new developments suggest the AI effort is taking shape.
He has bought 10,000 GPUs
- Tech companies use GPUs to train large AI models, given the computational workload required by newer technology.
- Musk’s AI initiative is in early stages, though the purchase of so much additional computational power shows he is “committed” to its development.
- He has criticized generative AI, saying the technology is powerful and needs regulation to ensure it operates within the public interest.
Twitter’s large language model
One of Musk’s new hires, Igor Babuschkin, said the project involves a large language model (LLM). An LLM is essentially AI that is trained on vast sets of data so it can eventually create new, sophisticated content and text on its own. Twitter’s massive amounts of data could train an LLM.
For instance, OpenAI, the creator of ChatGPT, previously had access to Twitter data for training purposes, though Musk said he put a stop to that in December.
He has hired AI talent
- By early March, he had officially hired the engineers Igor Babuschkin and Manuel Kroiss from Alphabet’s AI‑research subsidiary, DeepMind.
- Since at least February, Musk has approached people in the AI field with the idea of starting his own AI endeavor to rival ChatGPT.
Possible uses of Twitter’s generative AI
It is unclear exactly what Twitter will use generative AI for, but the people familiar said it could improve search or advertising. Musk has complained about search, and he even hired the security developer George Hotz for a three‑month internship to “fix” the search function. Hotz left after just one month.
Another possible use is advertising, the backbone of social‑media platforms. Twitter has struggled to woo advertisers since Musk took over and changed numerous policies. Generative AI has the capability, if trained for the use, to create new advertising images and text to target specific audiences.
Related stories
The information above is a paraphrased version of the original article and is intended for illustrative purposes.
Twitter’s GPU Upgrade: A Deep Dive into the Hidden Costs
Understanding the Financial Commitment
With Nvidia commanding roughly 95% of the GPU market, the new units intended for large AI workloads carry a price tag of $10,000. Musk’s Twitter has likely expended many tens of millions of dollars to support this massive upgrade, as reported by insiders.
Data‑Center Sites: Where the New GPUs Operate
- Atlanta – The primary location where the new GPUs will function.
- Oregon – A secondary data‑center site, still active for Twitter.
- Sacramento – Previously the main facility, operated by NTT, was shut down in late December after an abrupt decision by Musk.
Leadership Changes Driven by the Decision
Following the shutdown of Sacramento, engineering leader Behnam Rezaei resigned in response to the drastic move.
Musk’s Relationship with OpenAI and the Generative‑AI Landscape
Musk co‑founded OpenAI alongside Sam Altman and other tech leaders. After leaving the company in 2018, he has become critical of OpenAI’s direction this year, labeling it a “closed source, maximum‑profit” entity effectively controlled by Microsoft. Microsoft’s investment in OpenAI began in 2019 and expanded to a multibillion‑dollar partnership in January.
Contacting Insider for Insider Perspectives
Are you an insider or someone with relevant insights? Reach out to Kali Hays at khays@insider.com, on the secure messaging app Signal at 949‑280‑0267, or via Twitter DM at @hayskali. Please use a non‑work device for privacy.