Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
AI Apple

Apple Loses Fourth AI Researcher in a Month To Meta 22

Posted by msmash from the losing-faith dept.
Apple has lost its fourth AI researcher in a month to Meta [non-paywalled source], marking the latest setback to the iPhone maker's AI efforts. From a report: Bowen Zhang, a key multimodal AI researcher at Apple, left the company on Friday and is set to join Meta's recently formed superintelligence team, according to people familiar with the matter. Zhang was part of the Apple foundation models group, or AFM, which built the core technology behind the company's AI platform.

Meta previously lured away the leader of the team, Ruoming Pang, with a compensation package valued at more than $200 million, Bloomberg News has reported. Two other researchers from that group -- Tom Gunter and Mark Lee -- also recently joined Meta. AFM is made up of several dozen engineers and researchers across Cupertino, California, and New York. In response to the job offers from Meta and others, Apple has been marginally increasing the pay of its AFM staffers, whether or not they've threatened to leave, said the people, who asked not to be identified because the moves are private. Still, the pay levels pale in comparison with those of rivals.

Apple Loses Fourth AI Researcher in a Month To Meta More | Reply

Apple Loses Fourth AI Researcher in a Month To Meta

Comments Filter:

  • JFC (Score:1)

    by Anonymous Coward

    $200M? That's insane. Good for him I guess.

  • I don't want AI slop in every facet of my life. Siri absolutely needs improvement but I'd rather is not be able to answer a question than confidently give the wrong answer.

    • Re: (Score:2, Insightful)

      by larryjoe ( 135075 )

      I don't want AI slop in every facet of my life. Siri absolutely needs improvement but I'd rather is not be able to answer a question than confidently give the wrong answer.

      This is an interesting question. There are very few (maybe zero) systems that return confidence levels with answers, mainly because in most cases the combination muddles the results, making the results less useful. Aside from systems that essentially return the output of a function, e.g., like a calculator, pretty much all systems have less than 100% accuracy. So, anyone that requires 100% accuracy will be unable to use anything other than the calculator-type of system.

      I'm guessing most people have an ac

      • How I feel about wrong answers from AI is that they are like getting a search result from google that is irrelevant, or that is not credible and contains incorrect information.

        Neither search nor AI is perfect so you've definitely got to remain vigilant.

        I use search less and less as time goes by. ChatGPT gives you links to click on if you need to make sure (one could argue that if you're clicking the links it is indeed "search" after all). But I tend to use Google only if I need to. The AI they put i

      • I find that the occasional incorrect answer almost always either triggers my internal spider-sense or is a tangential or insignificant detail that I didn't need or won't use anyways

        How do you know that ChatGPT is correct the rest of the time?

        • I find that the occasional incorrect answer almost always either triggers my internal spider-sense or is a tangential or insignificant detail that I didn't need or won't use anyways

          How do you know that ChatGPT is correct the rest of the time?

          The same way I consider the accuracy of anything I read in a book, newspaper, magazine, Wikipedia, research paper, etc. I consider if it is consistent with what I already know and focus scrutiny on non-mundane claims and numbers. I focus on a small subset of numbers, claims, and details because most of the other stuff I don't care about. If that minor stuff is wrong, I don't care because it won't affect what I'm trying to do with the answer. However, the stuff I care about I consider skeptically.

    • It seems like that would be a better response to an announcement of a premature rollout, rather than to the hiring of researchers whose job it is to fix the problems you're talking about.
  • Because what better way to tell when something's gotten overblown than when Suckerberg starts spending tens of billions on it.

  • With $200M packages, they had to find the funds somewhere - no wonder they had all these layoffs recently.

  • Here's how this is going to go.
    "Hey, Meta, I need you to research applications for this compound I've come up with to fight cancer."
    "Sure, but first, let me show you a bunch of shit you don't give a rat's ass about. Then, I'll show you one result that might be relevant but I'll bury it quickly and show you more crap you don't care about."
    "Ugh, please sort this chronologically."
    "F*ck you. I'll decide what you get to see."
    "Dammit, I'm installing a third-party browser plugin to filter out your garbage."
    "F*ck

  • 200 million dollars compensation?
    He was a student researcher at google from 2020-2022 before becoming a director at Apple.
    41 research papers in ML, computer vision, signal processing.
    https://www.researchgate.net/p... [researchgate.net]
    Meta figures he can help them crack the AGI barrier so they can replace meat humans with virtual analogs.

    • 200 million dollars compensation?
      He was a student researcher at google from 2020-2022 before becoming a director at Apple.
      41 research papers in ML, computer vision, signal processing.
      https://www.researchgate.net/p... [researchgate.net]
      Meta figures he can help them crack the AGI barrier so they can replace meat humans with virtual analogs.

      If he is such a hotshot, then how come Apple isn't already much farther along with their AI Project?

  • But if you can get it...

    Pretty soon Apple won't have an AI team. They don't pay enough to have one, so they won't.

    • Maybe Apple can hire 20 very good researchers for $10,000,000 each instead of a single one for $200,000,000. I really doubt a single researcher is so good to justify the pay package. I understand that occasionally geniuses will do something that a team of 20 couldn't, but my bet is the probability that 20 very talented people are more likely to advance the art than a single person. But maybe I'm wrong - I don't have that kind of money to test my theory.

  • But it's not as if Apple couldn't afford to compete with Meta and others, if they chose to.

    • Tim Cook is an "operations" guy. I am sure he (and his team) carefully weighs cost / benefit. Apple has decided (rightly or wrongly) that it isn't worth paying such salaries. Mark Zuckerberg seems to be much less calculating - just look at how much he spent on his virtual world. He probably gets a kick out of "winning" by throwing around these huge pay packages whether it makes economic sense or not.

  • Wasn't that in American Gods?
    Neil, uh oh, Gaiman?

    This is not very believable to me. I have a hard time believing he is possibly worth it. You couldn't think like him and just get ... how many can you practically afford?

    Back to the question though, could this guy be bringing the special IP Sauce from his previous employer?

    For 200M, sure, I walk on water. It's gonna be worth it. Tap me.
  • Aside from this particular area, across the board, Apple has fallen behind in pay. Even the dinosaurs like AMAT and Lam are now competitive with Apple.
  • Meta? Isn't that the company that just a few years was going to have us all living in a virtual world? What happened to that? I don't think Zuke knows what he wants and he's just throwing stuff against the wall to see what sticks. Expensive stuff, but still, just stuff.

Slashdot Top Deals

"Now here's something you're really going to like!" -- Rocket J. Squirrel

Close