- Main tech corporations are battling for AI dominance, pouring tens of billions into infrastructure and providing sky-high compensation packages. Billionaire investor Mark Cuban notes this new section will see corporations locking down useful AI improvements and experience fairly than sharing them.
Main tech corporations are in a fierce competitors to be on the forefront of AI by recruiting high expertise, massively investing in analysis and improvement and infrastructure, and scooping up smaller startups left and proper.
This so-called AI warfare is about to turn into much more aggressive.
“What individuals are lacking about AI, [in my opinion], is not any firm goes to spend what’s going to find yourself being greater than a trillion {dollars} and settle for not being dominant,” Mark Cuban wrote Sunday on X. “They may discover a method to battle. I don’t know the way, apart from to guess it should get ugly.”
Tech giants like Meta, Google, Microsoft, and Apple have been ferociously recruiting high AI expertise—and doling out monumental pay packages to poach them. High-tier AI researchers at Meta are reportedly being supplied whole compensation packages of as much as $300 million over 4 years, with some first-year earnings over $100 million. One instance was a former Apple govt Ruoming Pang, who Bloomberg reported is becoming a member of Meta’s “Superintelligence” staff for $200 million.
In January, Microsoft introduced it was investing about $80 billion this yr alone on creating information facilities to coach AI fashions and deploy cloud-based functions. An AI researcher at Microsoft makes about $431,000, on common, in accordance with Ranges.fyi. A number of job-posting websites present the common wage for a software program engineer is roughly $125,000, so these so-called AI researchers are making greater than 3 times a software program engineer—a job that’s lengthy been thought to be one of the profitable in tech.
Google has additionally invested about $75 billion in AI this yr, and not too long ago introduced one other $25 billion funding to construct new AI information facilities. Nvidia can be a transparent frontrunner in {hardware} powering AI, and a few analysis scientists there reportedly make greater than $600,000.
“We’re seeing them rent away expertise and [intellectual property] to construct [and] advance their fashions,” Cuban wrote. “However that’s only the start.”
Cuban predicts main tech corporations can even begin shelling out extra to “lock up” the IP they assume their fashions have to preserve it from their rivals.
“The times of publish or perish are most likely over,” Cuban wrote. “It’s now publish, and its worth perishes, as a result of it received eaten up by each foundational mannequin.”
He suggests the businesses that develop useful IP for AI ought to encrypt it and silo it.
“IP is KING in an AI world,” Cuban added. “The occasions they’re a altering.”
