The artificial intelligence (AI) ecosystem faces two crucial bottlenecks around compute and research talent resources.
And the battle for both has never been fiercer, with nearly every tech company focused on building the most cutting-edge AI model vying for a small pool of engineers and scientists, as well as the GPUs and hardware plus cloud infrastructure necessary to actually build an AI system.
This, as Meta is intensifying the competition in the AI market by consolidating its two advanced AI divisions — the Fundamental AI Research (FAIR) team and its top-level Generative AI product team — into a single group.
“Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit. We’re bringing our two major AI research efforts (FAIR and GenAI) closer together to support this,” Meta Co-founder and CEO Mark Zuckerberg said in a recent Instagram post.
From an organizational point of view, this now means that Meta Chief AI Scientist Yann LeCun and FAIR VP Joelle Pineau will both now report to Chris Cox, Meta’s chief product officer, rather than the company’s CTO, Andrew Bosworth.
This strategic move underscores how Meta is now prioritizing product-level progress in developing general-purpose artificial intelligence chatbots and securing top talent in the competitive field of AI engineering — as opposed to attempting to lure top researchers to work on strategies like Meta’s metaverse, which is losing over $1 billion a month.
After all, the best researchers tend to want to work on the most ambitious problems. And while Zuckerberg’s attention was on the metaverse, his internal AI team was pushing out research breakthroughs, helping garner industry credibility with its open-source approach to AI development.
How Big Tech Is Defining the AI Landscape
Per Zuckerberg’s Instagram post, Meta is compiling “massive compute infrastructure to support our future roadmap” which includes almost 600,000 GPUs, a sizable war chest that is tied only with Microsoft’s projected stockpile and represents a compute capacity nearly three times larger than the next closest competitor.
But the race for supremacy in AI technologies has been quietly picking up speed over the past decade, and only bubbled to the foreground with the launch of OpenAI’s ChatGPT product.
Big Tech has been a big player in AI going back years. Over the past decade alone, Apple has made 32 AI acquisitions, Google has made 21, Meta 18 and Microsoft 17, according to data from CB Insights.
And the acquisition and investing spree is only continuing to accelerate as AI becomes the crown jewel of many tech company’s portfolios.
Microsoft, Amazon, Alphabet (Google’s parent company), and Nvidia raised around 8% of the total venture capital raised in North America during 2023, with over $25 billion, per Pitchbook.
That capital was deployed by the tech giants primarily across data and AI, with deals that represented just under a third of all the deal volume in the data and AI sectors.
Read also: Who Will Power the GenAI Operating System?
Google is a leading cloud provider and AI player, as are Microsoft and Amazon, meaning they are able to double-dip among their own services while also selling key services to other companies, including those they’ve invested in.
For example, compared to a traditional venture capital investor, if you are one of the Big Tech cloud providers and you invest $2 billion in an AI startup at a $20 billion valuation, and that company agrees to pay you several hundred million dollars a year to run their models on your infrastructure, you can still make money on the deal even if the investment itself does not work out down the line.
It is an attractive alternative model, and one that could see Big Tech entrench its hold over the AI landscape even further.
And as PYMNTS CEO Karen Webster wrote, one of the key trends to watch for 2024 will be the role of AI-native businesses in marginalizing incumbents, and as the competition for AI engineering talent and compute resources plays out, the impact it has on the AI marketplace will be one to watch.