3 ms·

3 years ago, I and many people were using GPT-3 on our iPhones and Apple Watches (see for example https://www.youtube.com/watch?v=tDB3uIEgU0E&t=36s https://www.

3 years ago, I and many people were using GPT-3 on our iPhones and Apple Watches (see for example https://www.youtube.com/watch?v=tDB3uIEgU0E&t=36s https://www.youtube.com/watch?v=tDB3uIEgU0E&t=36s). For 3 years I wondered why Apple wouldn't make it a native feature. - Was it because they didn't want to use a third-party AI system? No, because Apple Intelligence uses ChatGPT anyway. - Was it because Apple didn't know how to add AI features to iOS? No, if anything they have access to way more APIs than I had when I built the iOS Shortcut in the video. - Was it because Apple didn't take AI seriously? This to me sounds silly but in retrospect, I'm starting to believe this was exactly the case. Apple has been on a wrong trajectory for the past few years: - They were busy building Apple Car. Then when the AI pressure finally kicked in, they stopped the car project. Why were they in the car business in the first place?! And why did they simply tell most Apple Car employees to work on AI now? Cars also have AI but it's not the same thing. - They built Apple Vision Pro but have failed to showcase any real-world use case for it yet. Apple Intelligence isn't even supported on the AVP. I don't want to call it a failed project but by many measures it is. - Apple, and specifically Tim Cook, are great at optimization: from the supply chain to setting optimal prices that extract the most value from customers. But they forgot what allowed them to do this in the first place: a decade-long vision set by Steve Jobs. Current day Apple isn't going to disappear, but it's becoming less and less relevant in the broader tech landscape. In a sense, Apple has become the McDonald's of handheld gadgets. But if I want the latest tech, the most interesting use-cases of AI, or simply ground-braking devices no one has heard of, I'd go to a French/Persian/etc. restaurant instead of Apple's McDonalds.

vrighter 2y ago
Thinking AI is silly is not the wrong trajectory. AI IS silly. The law of large numbers applies. 99% accuracy means that thousands or millions of queries daily are just wrong. So AI is not suited for large scale deployments. When I was working with a company delivering a product to thousands (not millions) of customers, we still encountered very rare stuff like data corruption due to memory bit-flips. Quite regularly. Even though that's quite a rare event, much much less than 1% likely. A 1% failure rate for something deployed to millions of places is utterly abysmal. And no, they will not keep improving indefinitely. They've already hit a wall.
brookst 2y ago
Thanks for the obligatory “Apple is doomed” comment. I get it, and odds are one day it will be true so no harm trotting it out decade after decade, but I don’t think that time is now. - Watch and AirPods have both been incredible successful and were pure Cook-era Apple. Both were first envisioned after Jobs died. - Car was a failure, but it seems odd to criticize Cook for being a mere optimizer when he enthusiastically spent billions on the project. Which is it? No vision, or too-ambitious vision? - Agree that AVP has been disappointing. But again, so what? It’s similar to the Newton. Not every product will succeed. And again it’s hard to see AVP as a “McDonalds” effort. I don’t know what it is about Apple that attracts these “they are doomed if every product isn’t as revolutionary as the iPhone” takes. Who else is having world-changing home runs with every product?
- atonse 2y ago
  Totally agree with you and you didn’t even count their insanely great M chips that even have the interest of people running Linux and servers. Those chips are leading the industry in performance per watt and other metrics, and are fascinating as chips, not just Apple or ARM chips. Their laptops were “fixed” and many, including myself think they’re the best laptops we’ve ever owned. I feel the butterfly keyboard laptop era was when they were lost in the woods. But now they killing it. Even Apple Maps is great.
- behnamoh 2y ago
  Thanks for misinterpreting my comment. /s I said Apple is not gonna disappear (i.e., it's not doomed), but it's less and less relevant in the tech space. But you do you.

vrighter 2y ago

Thinking AI is silly is not the wrong trajectory. AI IS silly. The law of large numbers applies. 99% accuracy means that thousands or millions of queries daily are just wrong. So AI is not suited for large scale deployments. When I was working with a company delivering a product to thousands (not millions) of customers, we still encountered very rare stuff like data corruption due to memory bit-flips. Quite regularly. Even though that's quite a rare event, much much less than 1% likely. A 1% failure rate for something deployed to millions of places is utterly abysmal. And no, they will not keep improving indefinitely. They've already hit a wall.

brookst 2y ago

Thanks for the obligatory “Apple is doomed” comment. I get it, and odds are one day it will be true so no harm trotting it out decade after decade, but I don’t think that time is now. - Watch and AirPods have both been incredible successful and were pure Cook-era Apple. Both were first envisioned after Jobs died. - Car was a failure, but it seems odd to criticize Cook for being a mere optimizer when he enthusiastically spent billions on the project. Which is it? No vision, or too-ambitious vision? - Agree that AVP has been disappointing. But again, so what? It’s similar to the Newton. Not every product will succeed. And again it’s hard to see AVP as a “McDonalds” effort. I don’t know what it is about Apple that attracts these “they are doomed if every product isn’t as revolutionary as the iPhone” takes. Who else is having world-changing home runs with every product?

atonse 2y ago
Totally agree with you and you didn’t even count their insanely great M chips that even have the interest of people running Linux and servers. Those chips are leading the industry in performance per watt and other metrics, and are fascinating as chips, not just Apple or ARM chips. Their laptops were “fixed” and many, including myself think they’re the best laptops we’ve ever owned. I feel the butterfly keyboard laptop era was when they were lost in the woods. But now they killing it. Even Apple Maps is great.
behnamoh 2y ago
Thanks for misinterpreting my comment. /s I said Apple is not gonna disappear (i.e., it's not doomed), but it's less and less relevant in the tech space. But you do you.