Gruber is correct. Credibility matters more to Apple than many companies, especially since they market themselves as consumer privacy champions. If consumers don't trust Apple, it will be impossible to convince them that Apple is not selling their data.
The trap that Apple fell into is, making specific promises about AI capabilities is extremely difficult. Current AI is overhyped already, and Apple's privacy policies make it even more difficult to implement. "Autocorrect on steroids" is a good way to describe how AI currently works, and it works better when it has access to all your history.
So I'm not surprised that this AI marketing target failed, nor was I surprised when the Apple car failed. These failures are not unique to Apple - Musk and Uber also have failed spectacularly to deliver their big autonomous driving promises.
Apple still can make a ton of money on AI by delivering power efficient servers for it. And in fact they are planning to do that. They should leave the overpromising to more short-sighted companies, fire whoever painted them into this corner, and try to earn their credibility back.
The trap that Apple fell into is, making specific promises about AI capabilities is extremely difficult. Current AI is overhyped already, and Apple's privacy policies make it even more difficult to implement. "Autocorrect on steroids" is a good way to describe how AI currently works, and it works better when it has access to all your history.
So I'm not surprised that this AI marketing target failed, nor was I surprised when the Apple car failed. These failures are not unique to Apple - Musk and Uber also have failed spectacularly to deliver their big autonomous driving promises.
Apple still can make a ton of money on AI by delivering power efficient servers for it. And in fact they are planning to do that. They should leave the overpromising to more short-sighted companies, fire whoever painted them into this corner, and try to earn their credibility back.