tht

danox said:

tht said:

If it is an actual picture of the facility, it has solar PV over the parking lot. Love that! Hopefully they have tens of MWHr level batteries too. Solar over parking lots should be required all over the USA. Everything in the desert areas of Arizona needs to be painted white or high albedo as well. Not a pleasant place to be during the heat season.

I don't see this notion of owning this supply chain as viable anymore. So, the argument is moot.

30 years ago, the economics were such that it was possible. Development costs for fabs and customer bases were sufficiently low such that it could be done. Multiple micron-level fabs existed, that served that needs of a mainframe market or other small niche. Intel was ahead with leading edge fabs, but small companies could follow 6 to 12 months behind them with a near equivalent fab. AMD had its own fabs. Didn't Texas Instruments have its own fabs? TI! Before they gave up. IBM amazingly held on for a long while just serving their mainframe market.

Today, you need to sell into the entire population of the Earth to support the development of a leading edge, Angstrom level fab. It's Highlander rules, where there can only be one. It costs so much to develop a leading edge fab now, you need to have billions of chips shipped from your fab to make it worthwhile. To do that, you need to play a game of geo-politics, carving a path such that the chips coming out of the fab can be used everywhere in every nation.

If you don't do that, you will not get enough revenue to pay for the development of the fab. So, the notion that one nation can own the supply chain while serving the enter world is a fantasy. It's not even a fantasy. It's like saying gravity is optional or energy isn't conserved. The economics aren't going to work out.

Intel, IBM, Xerox, Motorola Schaumburg, Illinois, and US Steel failed due to management changes, the inability to recognize the importance of long-term product iteration and the need for ongoing research and development.

One other problem in America is the blame game, where production workers are blamed despite management’s bad decisions, which is another on going long term issue.

In America, MBAs, accountants, financial/bankers, and lawyers often end up running companies, especially in specialized fields like technology, pharmaceuticals, and high-end manufacturing. However, these elements should not be allowed to make strategic decisions or to run such companies, support yes run the company. Hell no….

Another big element is the time it’s gonna take to turn things around and no one in America wants to hear it but it’s probably overall two decades or more which is going to be a big problem in America because that’s stretched over five administrations oh boy……

Is there really a problem? The USA has 4% unemployment, has trouble filling manufacturing jobs, and has one of the highest GDPs among nations in the world? Right? What's missing? What are you sensing that is wrong?

Yes, the incentive structures of public companies can lead to some bad outcomes for companies. People have been aware of it for a very long time. Very successful enterprises can be ruined by just a few people, like Boeing or Intel. I don't think you can fix it as assholes are an inherent part of humanity. Some companies will have a spiral of bad decisions they can't recover no matter the structures erected to prevent it. There's probably government incentives to make businesses easier to capitalize, easier to build up to replace failed companies though.

Then, hard to believe a lot of these incumbent companies can maintain. Like Japanese car companies have essentially ceded, voluntarily, the car market to Korean and Chinese car companies, and whoever else up and coming. They won't be able to catch up. Those decisions are ingrained in some incumbents and it takes special leadership to effectively burn the ships and move forward. There needs to be incentives to make it easier for new businesses to pop up. And this is assuming the gov't knows what to do, and a lot don't.

If it is an actual picture of the facility, it has solar PV over the parking lot. Love that! Hopefully they have tens of MWHr level batteries too. Solar over parking lots should be required all over the USA. Everything in the desert areas of Arizona needs to be painted white or high albedo as well. Not a pleasant place to be during the heat season.

I don't see this notion of owning this supply chain as viable anymore. So, the argument is moot.

30 years ago, the economics were such that it was possible. Development costs for fabs and customer bases were sufficiently low such that it could be done. Multiple micron-level fabs existed, that served that needs of a mainframe market or other small niche. Intel was ahead with leading edge fabs, but small companies could follow 6 to 12 months behind them with a near equivalent fab. AMD had its own fabs. Didn't Texas Instruments have its own fabs? TI! Before they gave up. IBM amazingly held on for a long while just serving their mainframe market.

Today, you need to sell into the entire population of the Earth to support the development of a leading edge, Angstrom level fab. It's Highlander rules, where there can only be one. It costs so much to develop a leading edge fab now, you need to have billions of chips shipped from your fab to make it worthwhile. To do that, you need to play a game of geo-politics, carving a path such that the chips coming out of the fab can be used everywhere in every nation.

If you don't do that, you will not get enough revenue to pay for the development of the fab. So, the notion that one nation can own the supply chain while serving the enter world is a fantasy. It's not even a fantasy. It's like saying gravity is optional or energy isn't conserved. The economics aren't going to work out.

hammeroftruth said:

The big problem is the idea of a chatbot Siri sounds great, but when you factor in Apple’s fundamental pillar of privacy, it makes developing a product exponentially harder and takes a lot more time. 

The other complications are integrating Siri with other platforms like HomeKit and having it smart enough to know you, your smart devices at home, and your other Apple devices and work together while keeping your privacy. 

Sure, you can have your data on your device which is nice when it comes to privacy, but I haven’t been doing much actively using Apple intelligence and the database is 5gb. What happens in a year, especially if I start using it daily? They’re gonna have to figure out how to sell us personal iCloud storage for AI that is secure so nobody, even Apple, can access it or disseminate it. This is something I see happening within 6 years. 

Is it really a great idea, though? The narrative around AI is akin to it being a nuclear weapon or an "iPhone" in 2007. Lots of rah rah with it, but is it really beneficial in Apple's context?

What do people want a LLM chatbot to do? Is what they want to do something a LLM chatbot can do? What can it do on a phone?

1. LLM services are great as advanced search services as they complete the next step for what you want to do. It is a great feature for professional workflows, both as a service and implementation tool, which you don't do on a phone. Apple really doesn't offer this in its portfolio. They have done perfectly well without offering a search service (a la Google), a social media service (Facebook, etc), and whatever is out there. Asking Apple to offer search services is a very similar question, and they have been fine without it. How are the circumstances different?

So, I think they can just let this percolate among the community of LLM developers as they do their own. They don't offer any services open to everyone, save for Apple Music. Anything else? Not in their wheelhouse, nor is it needed for them to be successful.

2. What can an LLM chatbot do for you on a phone? A consumer features question.

Have been thinking about this for a while, and I don't have anything great. Summaries of web pages? It would be nice as virtually every web page is optimized to make the article as long as possible in order to maximize ads. LLM summaries will eventually be gamed to uselessness, possibly even outlawed if ads are stripped out. The gamification will surely result in the LLM summary containing the ads too.

An LLM service that is able to filter out or identify LLM content will surely be offered to help people avoid the chaff, just like there are email filters, message filters, and whatnot. We actually need something like this for human generated content too. An LLM service that reformats web pages to make them readable and reduce power consumption? Yet another ads and scams are destroying everything take there.

How about arranging a trip somewhere? I can't see how this will effectively work, unless Apple, or LLM service, owns all the parts including credit cards. Like, the prompt will take just as much work to write as searching and buying yourself. Then, which service would like to participate? Expedia? The bank? The airline? 

CarmB said:

How many consumers have expressed dismay over the thickness of current iPhone models? Why is making the iPhone thinner a thing?

Apple wants to sell iPhones. They think they will get enough buyers for the iPhone 17 Air, at about $900 base price (?), to make it a worthwhile effort. There is a niche of people who want a large display phone that is light and thin. It's basically that simple.

There's also a niche of people who want a small model, about 5.5" display, too. I'd like Apple to offer that too. Like I said before, it is a margins game. A 5.5" model that has modern components will need to cost $800 to be worthwhile, otherwise, either Apple uses prior year components to price it at $700, or do what they have done, which is not to offer it. I do think that an iPhone mini with n-1, or n-2, components would be worth it, but I'm not Apple.

It is likely that Apple will have better margins on a $900 iPhone 17 Air versus a $700 iPhone mini. They keep trying at that $900 price knowing that, at least. An $800 iPhone mini would be a tough sell.

You should also note that the nominal iPhones, with about 8 mm of thickness, have a stacked logic board system. Two PCBs are layered on top of each other, and hold all the computing chips, resistors, capacitors, ports for connectors etc.

With the 5.5 iPhone 17 Air, this two layer board system likely can't be employed. This will make the footprint for its logic board larger relative to the 6.7" models. So, they likely only could use 1 camera, possibly one speaker, and the battery still has to be smaller than a battery of the same footprint would imply. Taptic engine could get smaller. Not much else is left. Well, no Face ID is an option.

But, if it is as light as a 4.7" iPhone SE, that's a big usability win.