I’ve refrained from commenting on any of these posts bc I know everyone has something to say, but this still feels quite shocking even if the project stagnated like almost nothing I’ve seen before in my Apple-nerd days. $10B is an inconceivable amount of money—a massive expense even for the most valuable company on the planet. I wonder if they got stuck in the sunk cost fallacy with this one.
Apple was more so likely chasing 'lottery tickets' more so than caught up in 'sunk costs'. Pragmatically there was no costs. Apple was making at least $1B per year on revenue sharing with Google on Apple end user activity. The users do all the 'work' and the 'free' money just flows into Apple. The users do predictably the same amount of work every year so there is no cost sink here because it just get covered by a steady flow from
Apple's product lines are all profitable, so they can pay for their own R&D cost overheads. This is just 'free' money to gamble with every year. Also help to stick it at a project that Google is trying to fund so that Goolge keeps spending more money on that than on something that would be more competitive with Apple's profitable products. (e.g., Google spend more money on Waymo than on making their own SoCs for smartphones. Apple uses part of Google's own revenue flow to do that; so it isn't 'costing' the rest of Apple any money to tweak Google into that investment allocation. )
As Google's payments went from $1B to $2B to $3B to $4B, it just gets easier and easier to toss $1B at 'lottery tickets'. Furthermore, it gets easier and easier to hide the boondoggle on the books.
Only reinforced by Telsa's stock going off into stratosphere. If Apple managed to get lucky and managed to pull something off here better than Google/Tesla/etc it would pay off. If it doesn't... then it is just 'fun' to play the lottery 'extra' cash.
If suspect it is not a coincidence that Apple kills the Car project about the same time:
1. they go to the DoJ and beg to avoid antitrust.
Apple representatives met with the U.S. Justice Department last week in a last ditch effort to persuade the agency not to file an antitrust suit...
www.macrumors.com
2. they Govt has basically outed their 'effort free cash flow from users' deal with Google.
3. The EU is poking around at their 30% AppStore tax.
If the billion dollar 'free' money from Google stops then Apple would have substantially less lottery ticket buying money.
All the folks gyrating about how $10B is 'a lot of money, how could Apple spend that much and not miss it'. How much money got bet on the Super Bowl a couple of weeks ago? How much money has USA PowerBall and MegaMillions collected in the last 10 years?
Even if the billion dollar 'free' money doesn't disappear from Google, "AI" is a far more trendy lottery ticket to buy now. The EV car market has lots of competitors now. It is unlikely Tesla is going to win the "take over the world" EV wars ( not the #1 selling brand and likely only going to get smaller share going forward. )
Plus, "AI" has way more synergy with the rest of the product lines. End users might actually get some indirect benefit for all those billions they are generating for Apple for 'free'.
The spatial reasoning that the R1 does and some of the basic spatial reasoning that a car would need to do has some overlap. So likely there is some narrow sharing there ( so it isn't $10B completely flushed down the drain. ). The 'free' money allowed super loose coupling. If the 'free' money disappears then the coupling would have to get much tighter.
I do dream of truly autonomous driving for personal reasons—alas, I don’t think it’s realistic, barring some unforeseen technological breakthrough. (such breakthroughs are kinda the nature of R&D, but…yknow)
Over the long term it is realistic. However, as a 49'er gold rush contest to be first to claim success not as much. The testing/regulation rigor is relatively weak for the high bar that should be set. And there is too much "if we just pile up a big enough pile of data the answer will magically pop out the other side" hysteria. AI has gone through lots of cycles of overpromise and underdeliver. But if measure things over decades there is substantive progress.
The huge folly of the Apple car was primarily designing around what they wished they had as opposed to what they had.