Elon Musk reportedly buys tons of GPUs to cook up AI masterplan with Twitter

Twitter is just embarking on a big AI project with significant ramifications for the future, according to a new report that reckons Elon Musk has some serious plans underway to take on the ChatGPT juggernaut.

This comes from Business Insider, which puts forward the claim that Musk has already bought around 10,000 GPUs, as well as having recruited a pair of AI experts (from DeepMind, an Alphabet outfit now involved in the development of the Google Bard AI).

The idea is to build an AI of the large language model (LLM) variety like ChatGPT, but this venture is still very much in the early days for Twitter, according to one of the sources tapped by Business Insider.

Clearly, though, Musk is serious enough about pursuing this avenue of development if he really has forked out a pile of cash for thousands upon thousands of graphics cards (presumably very expensive, heavyweight GPUs at that) to drive the effort.

To what end?

It's one thing to talk about Musk possibly knocking up his own AI, and another to guess exactly what this entity might do for Twitter. It seems an obvious bet that the project will be aiming along the lines of better monetizing the platform, perhaps through more accurate targeting of advertising?

Or there may well be multiple goals at work, on the one hand with the idea of helping Twitter users on the face of it, and at the same time, doubtless finding a route to make some cash into the bargain.

It's likely Musk will be on the hunt for more top AI engineers, we're told, and that this recruitment drive (which has picked up at least two staff, as mentioned) has been ongoing since February.

All this apparently happening is noteworthy against the backdrop of a recent open letter signed by Musk and other big names from the tech world (like Apple co-founder Steve Wozniak) warning about the fast progress of AI and its potential threat to society.

The more cynical could get pretty skeptical about Musk chiming in with the call to put the brakes on with AI being more about letting him catch up, than it is about concerns regarding the wider impact of AI on society and humankind.

Naturally, whatever the agenda for Musk's apparent plans for AI domination turns out to be, there are bound to be worries about the privacy of user data, and how any artificial intelligence system might use that data.

Privacy is a matter that has been brightly focused under the media spotlight recently after we saw Twitter Circles go badly wrong. Indeed, users ended up having nude photographs (that were supposed to be kept secret within the secure circle) publicly aired, which was embarrassing for them, and certainly for Twitter as well.