3 ms·

Who is this best of the best that they didn't hire?

m0zg 8y ago
Everyone who cares even a little about their academic career. Apple doesn't really let you publish. Even pop-sci articles they do publish on their website don't have any names. I know no sane researcher who would agree to end their academic participation so abruptly, particularly in order to work for an organization widely perceived as a perennial laggard.
- dwighttk 8y ago
  so is it "they won't open their wallet!" or "they won't let you publish!"
  - m0zg 8y ago
    How about both. Google, Microsoft, and Facebook both pay their researchers very well and let them publish extensively. That's why they have world-leading research organizations, and Apple does not.
    - danso 8y ago
      If Google’s strategy is to be emulated, then why is hiring someone who was part of that strategy a bad move on Apple’s part?
      - m0zg 8y ago
        I'm not saying it's a _bad_ move. It's just insufficient. And until they unclench wrt publishing nothing is going to change for them. And even after they do that, it will take a lot of convincing (and large wads of cash) to hire the best talent that can not only get Siri to catch up to Google, but overtake it. My problem is not with Giannandrea's new position, it's with the lack of urgency on Apple's part. My patience has been wearing thin as of late.
        tibyat 8y ago
        you’re the mark jackson of tech commentating. you refuse to commit to any fact, likely because you dont have any to share, yet you feel oh so strongly for your argument, so you try to base it on intangibles like “lack of urgency”.
    - psychometry 8y ago
      Do you have evidence that Apple pays AI researchers less than those other companies? Honestly I'm not sure that you even understand the argument you're trying to make. The rest of us here certainly don't.
      - woolvalley 8y ago
        In my experience, the offers that google, fb & netflix give are better than the offers that amazon, msft & apple give on a consistent basis. 'Special' people will get special offers, but we would have to compare their offers of what they would get elsewhere, and if their special skill would be equally valued at other companies. Apple has an institutional memory of almost dying, so they can be a very 'cheap' company under the hood when they can get away with it. It reflects in their pay.
      - bad_good_guy 8y ago
        Just because you aren't understanding doesn't mean most others aren't. He clearly said the other companies have an AND situation in regards to pay and publishing. He is saying Apple does not, as it does not allow proper publishing. I.e. {'Pay' AND 'Publishing' > 'Pay'} not {Apple 'Pay' < Other 'Pay'}
- rrdharan 8y ago
  Actually they fixed that 2 years ago: https://www.theverge.com/platform/amp/2016/12/6/13858354/apple-publishing-ai-research-siri-self-driving-cars https://www.theverge.com/platform/amp/2016/12/6/13858354/app... https://machinelearning.apple.com/ https://machinelearning.apple.com/
  - m0zg 8y ago
    But... they didn't. In the very page you linked, try to find a single name of a concrete researcher who wrote/contributed to the article. If there's no name on the paper it's not "publishing".
    - kenferry 8y ago
      https://9to5mac.com/2017/08/25/apple-wins-best-paper-prestigious-conference/ https://9to5mac.com/2017/08/25/apple-wins-best-paper-prestig... Names are there. Check the arxiv link.
    - Q6T46nT668w6i3m 8y ago
      They won the Best Paper CVPR prize in 2017. Learning from Simulated and Unsupervised Images through Adversarial Training by Ashish Shrivastava, Tomas Pfister, Oncel Tuzel, Joshua Susskind, Wenda Wang, & Russell Webb.

m0zg 8y ago

Everyone who cares even a little about their academic career. Apple doesn't really let you publish. Even pop-sci articles they do publish on their website don't have any names. I know no sane researcher who would agree to end their academic participation so abruptly, particularly in order to work for an organization widely perceived as a perennial laggard.

dwighttk 8y ago
so is it "they won't open their wallet!" or "they won't let you publish!"
- m0zg 8y ago
  How about both. Google, Microsoft, and Facebook both pay their researchers very well and let them publish extensively. That's why they have world-leading research organizations, and Apple does not.
  - danso 8y ago
    If Google’s strategy is to be emulated, then why is hiring someone who was part of that strategy a bad move on Apple’s part?
    - m0zg 8y ago
      I'm not saying it's a _bad_ move. It's just insufficient. And until they unclench wrt publishing nothing is going to change for them. And even after they do that, it will take a lot of convincing (and large wads of cash) to hire the best talent that can not only get Siri to catch up to Google, but overtake it. My problem is not with Giannandrea's new position, it's with the lack of urgency on Apple's part. My patience has been wearing thin as of late.
      - tibyat 8y ago
        you’re the mark jackson of tech commentating. you refuse to commit to any fact, likely because you dont have any to share, yet you feel oh so strongly for your argument, so you try to base it on intangibles like “lack of urgency”.
  - psychometry 8y ago
    Do you have evidence that Apple pays AI researchers less than those other companies? Honestly I'm not sure that you even understand the argument you're trying to make. The rest of us here certainly don't.
    - woolvalley 8y ago
      In my experience, the offers that google, fb & netflix give are better than the offers that amazon, msft & apple give on a consistent basis. 'Special' people will get special offers, but we would have to compare their offers of what they would get elsewhere, and if their special skill would be equally valued at other companies. Apple has an institutional memory of almost dying, so they can be a very 'cheap' company under the hood when they can get away with it. It reflects in their pay.
    - bad_good_guy 8y ago
      Just because you aren't understanding doesn't mean most others aren't. He clearly said the other companies have an AND situation in regards to pay and publishing. He is saying Apple does not, as it does not allow proper publishing. I.e. {'Pay' AND 'Publishing' > 'Pay'} not {Apple 'Pay' < Other 'Pay'}
rrdharan 8y ago
Actually they fixed that 2 years ago: https://www.theverge.com/platform/amp/2016/12/6/13858354/apple-publishing-ai-research-siri-self-driving-cars https://www.theverge.com/platform/amp/2016/12/6/13858354/app... https://machinelearning.apple.com/ https://machinelearning.apple.com/
- m0zg 8y ago
  But... they didn't. In the very page you linked, try to find a single name of a concrete researcher who wrote/contributed to the article. If there's no name on the paper it's not "publishing".
  - kenferry 8y ago
    https://9to5mac.com/2017/08/25/apple-wins-best-paper-prestigious-conference/ https://9to5mac.com/2017/08/25/apple-wins-best-paper-prestig... Names are there. Check the arxiv link.
  - Q6T46nT668w6i3m 8y ago
    They won the Best Paper CVPR prize in 2017. Learning from Simulated and Unsupervised Images through Adversarial Training by Ashish Shrivastava, Tomas Pfister, Oncel Tuzel, Joshua Susskind, Wenda Wang, & Russell Webb.