I sure hope we eventually have the ability to run a local model at home (think a limited version of SARAH from the show Eureka, though maybe without Fargo's voice). And it could interface with our mobile devices over a VPN. Imagine pairing that with augmented glasses. We'd still have to use commercial APIs for a lot of stuff, of course, but we wouldn't have to worry about porting our data over if we switched providers.In about 5 years, we will all subscribe to a personal AI that will slip in and out of all devices with APIs from the host company. We will get embedded with the personal AI so switching will be difficult.
Apple really screwed things up here in terms of development. All the key players went to Google and then they did this.
Yes. I wanted to keep it shortPerplexity uses models from various companies.
Interesting twist. Initially Apple seemed to have been going in the OpenAI direction. Seems like the existing multi-billion dollar relationship they have with Google prevailed over starting something new with another company? Obviously Google had to demonstrate that their AI is as competent as whatever OpenAI was bringing to the table.
What does any of that have to do with this topic?yea goodbye siri. I'll build my own local setup with home assistant and start migrating off the apple ecosystem. If they are giving up, so am I.
Why would it matter though - once they buy access to the LLM and run it on their own servers, they wouldn't be relying on OpenAI, Anthropic or Google.I think there is also the stability factor. If this AI bonanza starts to go sideways, OpenAI is fully exposed, whereas Google has something else to fall back on to keep things going.
This is a joke. I try to use as little services from Google as possible.
I would not trust anything that's being transmitted from Apple products to Siri from now on.
How can you say it's a "win" for Apple and then say in the same breath "it's clear that Apple isn't capable of doing this themselves" lollllllllllllllIf this is Apple simply licensing a Google model and running it privately on their own servers (without transmitting data to Google), this should only be seen as a win. It's clear that Apple cannot make their own foundational model, so licensing arguably the strongest existing model from a third party is the next best move.
Will the 1 trillion parameter model finally give Siri enough horsepower to figure out how to level the volume with other apps? If I'm let's say streaming YouTube on a browsers and ask Siri something, it pretty much yells at me. Might want to fix that first.
I have had Apple devices(Mac, iPad, and iPhone) since 2003 and I have never been able to use Siri. I will knowingly use AI if it makes it useful.I have had Apple devices(Mac, iPad, and iPhone) since 2013 and I have never used Siri. I will never knowingly use AI.
LOL. Been listening to this same broken record about Apple for decades and they're still here.Beginning of end. Apple stumbling.
They should just buy Perplexity and catch up to the Assistant/AI features Android and others have had for a while now.
Edit: Didn't realize Perplexity was just a wrapper, although it looks like they are also working on their own LLM. Anthropic would be an option and I'd argue a "buy once cry once" mentality, especially given how much cash Apple has on hand. They are falling behind fast and it's unfortunate how much turnover they have had, mostly thanks to all the poaching. I know Apple is never first to the party but at this point their AI footprint is almost non-existent and I'd like to hope they catch up.
If Apple announced this today how credible is it their own LLM would be ready in 2026?
It's honestly disgusting at this point. I have disabled Siri on every single Apple device I've owned for years because I have no interest in talking to my devices when I can do things faster myself without speaking aloud to the room. The whole AI bubble is about to burst as most people realize it's all fake anyway. And now Apple will just jack up the prices of all their devices again to compensate for this huge Google bill they will be paying out.Apple is now dry sucking to Google. Freakin' sellout decision and AI is overhyped. Wish Scott Forstall took over instead of dismissing him.
Yes, you are right and I am not sure what the scale of it is but I am sure it is not outlandish what everyone else is doing. Interesting to hear more about their chips. Good to see they have their own solution and are not relying as much in on Nvidia like everyone else. And not having their Ai data center up and running could be a huge reason or at least partly why Apple has not rolled out it's revised AI solution.Apple IS building its own big data centers, containing its own Apple-manufactured servers, on which it will run Google's AI LLM. Its Apple's "Private Cloud Compute" initiative. Those data centers may not need to be as massive as some others, since Apple isn't planning on including a general, universal chatbot capability for Siri/Apple Intelligence, though I wouldn't be surprised if they change their mind about that at some point.
so I can say... Siri... is it raining outside right now?
The smarter, more capable version of Siri that Apple is developing will be powered by Google Gemini, reports Bloomberg. Apple will pay Google approximately $1 billion per year for a 1.2 trillion parameter artificial intelligence model that was developed by Google.
![]()
Siri will be able to answer more complex queries and complete more complicated tasks in and between apps. It will be closer in function to Claude and ChatGPT, though Apple is not planning a dedicated chatbot app.
Article Link: Apple's New Siri Will Be Powered By Google Gemini
apple will only do so if they can monetize it... they do not want a loss leader... has anyone made real $$$ from AI other than the chip makers???Yes, you are right and I am not sure what the scale of it is but I am sure it is not outlandish what everyone else is doing. Interesting to hear more about their chips. Good to see they have their own solution and are not relying as much in on Nvidia like everyone else.