Apple signs deal with OpenAI for iOS, still wants Google as an 'option'

This is a pretty strongly worded lecture when you’re talking about something that hasn’t even happened yet.

But doesn’t Apple make sure things break when specific processes cannot reach specific server?

Are we absolutely positive that the cables for that switch is connected somewhere, anywhere*?

*: Except a little LED which glows when you throw the switch.

> Isn't Siri required for speech processing?

Probably, but I don't use it, only use my car buttons/dial for controlling CarPlay. The times I've tried the voice control it got maybe 50% of the instructions correct, which was more distracting than the muscle-learned buttons/dial control, so I keep using what works.

If I was trying to use voice control, it'd make sense. But I don't, so it doesn't.

Forcing me into a voice assistant enabled at all times regardless though kills the whole thing. I end up preferring buttons and gui because the hands-free "assistance" is most often exercises in frustration and I conclude not worth it.

At this point many years of poor quality experience whenever I encounter voice processing is the matter. Happy to call me old, stupid, and bad at speech ... all are probably true. Still won't fix the frustrations I have for Siri, Tesla, Google-something, Bixby's button, and Alexa devices ... much less every customer service line now consistently apologizing that they didn't get that.

Carplay is leaps and bounds a better approach to gui/console in most cars. Much better than Subaru, Ford, Toyota or others can produce and there shouldn't be a voice assistant cost along with it.

Google is as bad or worse, but there are plenty of other choices. Run a FOSS tablet, or forgo using a phone at all. They really are optional. I have not carried a phone in years.

Apple offers people an extended social graph, support, and great UX so long as they leave their freedoms at the door of the walled garden. They win people over with carefully crafted psychological manipulation.

This is toxic behavior that stunts societal progress for money.

That is like a cult in that it exploits peoples tribal nature to lock them in to an ecosystem for the gain of those at the top.

I find that to be evil. I also find Microsoft, Google, and Meta to be just as bad if not worse, as well as all companies who emulate their behavior.

We should be selling users services, useful tools without restrictions, increased freedom, increased privacy, and increased defense against centralized entities that may put them at risk for profit or power.

The Lumia phones were great. The first iterations of the OS tried to integrate 3rd party services deeply into the OS. (News, Social, Mail, Calendar aggregations) But the service-providers did not like that because: no advertising possibilities.

Then they had to shift towards apps, but at that time they were way too late.

If ARM matures, I could see a new Windows Phone.

Contribute to open source projects.

Same as always, apply. Specifically with OS stuff you’ll need to show an aptitude for it, easiest way is if you’re already a college grad and you took an OS design track. If you’re self taught, it’s much harder to prove you have the basics down, so as another person suggested, contribute to OSS projects, which also sounds daunting, but unlike web dev, OS dev is not somewhere where it’s ok to be mediocre.

> I think that Google & Apple's trust relationship has broken down post-Google Maps.

I think Apple's "trust" can be bought by a big enough customer, and Apple's continued interaction with Google's privacy-destroying search engine seems to demonstrate that nicely.

Apple doesn't typically offer options to the user like this. They'll decide if what's best for the given context.

I suspect that current ML tech is good enough to constitute a big discrete step forward in terms of how much value Siri is providing as a digital assistant that's also in charge of your home. If that's the case, there's more harm in being reputed as the company that is one serious step behind in a space that isn't yet settled vs facing the same level of gaffs that Google and Microsoft are putting out.

Just look at Microsoft right now. They're experiencing all the heat that OpenAI brings, including unforced errors like Sky. Reputationally speaking, are they better or worse off having made such an early push for LLM product integration?

It can't be any worse than how bad Siri and Alexa are right now. Amazon gave me one of those Pop things recently and it was set up with the wrong zip code. I kept asking it for the weather in Chicago, every way that I could think of, but it had no idea what I was trying to ask for. The speaker is great in it, though.

> Hallucinations would be terrible

Tighten the temperature.

Everyone is trying to sell AI as AGI. Apple’s genius will be in breaking from that. Simply, dependable, tightly controlled LLM that isn’t even branded as AI. Just a thing that does certain tasks for you better.

They could use OpenAI only as the natural language processing part, and constrain it to only using the vetted services/actions that Siri already can do rather than answering open-ended questions. That would still make Siri massively more useful than it is today since it would remove the need to word things in a specific way.

It wouldn't be bullet-proof (there are always "jailbreaks" to these kinds of prompts) but it might be good enough if they slap a "beta" tag on it.

I went back to iOS recently, and with AirPods or CarPlay, I found Siri to be a really pleasant surprise. If I need something done, like a text message sent or a quick calculation done, I can just ask without taking my hands or eyes off what I’m doing. It’s also gotten a lot better at taking the hot word + command + arguments without breaks. “hey siri, send a message to dan I’m here” gets you “to dan: I’m here, send it?” For CarPlay, “hey siri, navigate to xyz restaurant” works great to start the navigation if you change your mind about where you’re going once you’re already on the road.

Combine that with “announce notifications” and it really feels like the future, “dan says, are you on your way” and it listens for a few seconds, where you can just say “reply yep I’m 15 minutes away”.

They’re already bringing AI into it a bit, if you get a picture it will try to describe what’s happening in the picture “dan sent a photo of a piece of paper with text on it” which is interesting but it’s not getting enough detail in the descriptions to be of much use just yet. A good multimodal model could definitely improve that, though it must be noted I prefer a worse on-device model to a server based model that can collect a bunch of data.

I think for really off the wall questions/brainstorming, going out to an LLM would yield better results than the current behavior of pointing you towards Google in the browser. “hey siri, what are the typical speeds people achieve on a SPI bus” would be nice if it could give some answer even if it’s not always 100% perfect, it’s kinda like bouncing questions off a slightly unreliable coworker, but you know, better than nothing when you can do it hands/eyes free.

I also have a hope that AI (although maybe not specifically LLMs) would help these assistants handle more complex tasks, like “add eggs to my grocery list” where your grocery list is just in a note and not in some special place. It doesn’t feel out of the question given current technology, I just think nobody’s done it yet.

> You can't trust the LLMs, so you can't use them for anything where you aren't going to go through the output meticulously afterwards.

People keep saying things like this, and I’m confused because I go to Google for facts, and chatgpt is still super useful. Things I’ve used chatgpt for recently:

- Here’s some text I’m writing for a paper. Help me reword it to make it sound more natural.

- I want to put coconut crème into a stew. When should I add it? What else would go well?

- I’m making a Makefile which has some unusual requirements. How do I do that? Here’s the actual commands I want it to run. Write the makefile. (The makefile wasn’t great, but it was a good start and let me hit the ground running.)

- Generate 100 quirky characteristics for an npc in D&D

And so on. It’s a tool for creativity. I absolutely can’t wait for my HomePods to be replaced with chatgpt.

But people need a new phone eventually. If you continuously manufacture phones anyways you might as well update the hardware while you're at it. (It also helps with keeping the price up at >1k)

I for one wouldn't buy a one year old new phone. Used sure but then the price is right.

> They also need to make 4 inch phone, like iPhone SE 1. That was the perfect size for a phone.

There are dozens of people who agree with you ;)

Apple came late to the big phone and were pulled there by the market. I'm surprised they still release a smaller phone at all given the rumors of such low sales.

They would be giving up massive amounts of profit if they did that. There would be a serious case for investors to sue if they made that decision.