Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Apple picks Gemini to power Siri

cnbc.com

Post reply on HN
Print view View original

51–60 of 674 posts

Re: Apple picks Gemini to power Siri

#51

Post by mathieuh »

drcongo wrote: post #38
johnthuss wrote: post #23

Earlier quoted context omitted.

The Google Maps situation is a great example of why this will be hard. When Apple switched to their own maps it was a huge failure resulting in a rare public apology from the company. In order to switch you have to be able to do absolutely everything that the previous solution offered without loss of quality. Given Google's competence in AI development that will be a high bar to meet.

I was in agreement with the parent before I read this, and now I'm in agreement with you. It is a great example, I know so many people who never switched back to Apple Maps because it was so poor initially. Personally I find it a considerably better experience than Google Maps these days, but those lost users still aren't coming back.

Mobile digital mapping was already a useful thing though. Even though Apple Maps was initially a failure I still came back to it every so often to see how it was progressing and eventually it ended up pretty good.

Maybe I'm weird but mobile assistants have never been useful for me. I tried Siri a couple of times and it didn't work. I haven't tried it since because even if it worked perfectly I'm not sure I'd have any use for it.

I see it more like the Vision Pro. Doesn't matter how good the product ends up being, I just don't think it's something most people are going to have a use for.

As far as I'm concerned no one has proved the utility of these mobile assistants yet.

Re: Apple picks Gemini to power Siri

#52

Post by ysnp »

Fiveplus wrote: post #42

The writing was on the wall the moment Apple stopped trying to buy their way into the server-side training game like what three years ago? Apple has the best edge inference silicon in the world (neural engine), but they have effectively zero presence in a training datacenter. They simply do not have the TPU pods or the H100 clusters to train a frontier model like Gemini 2.5 or 3.0 from scratch without burning 10 year…

Could you elaborate a bit on why you've judged it as privacy theatre? I'm skeptical but uninformed, and I believe Mullvad are taking a similar approach.

Re: Apple picks Gemini to power Siri

#53

Post by drob518 »

Fiveplus wrote: post #42

The writing was on the wall the moment Apple stopped trying to buy their way into the server-side training game like what three years ago? Apple has the best edge inference silicon in the world (neural engine), but they have effectively zero presence in a training datacenter. They simply do not have the TPU pods or the H100 clusters to train a frontier model like Gemini 2.5 or 3.0 from scratch without burning 10 year…

Yea, I think it’s smart, too. There are multiple companies who have spent a fortune on training and are going to be increasingly interested in (desperate to?) see a return from it. Apple can choose the best of the bunch, pay less than they would have to to build it themselves, and swap to a new one if someone produces another breakthrough.

Re: Apple picks Gemini to power Siri

#54

Post by burnte »

johnthuss wrote: post #23
Angostura wrote: post #16

Earlier quoted context omitted.

I don't see why - iOS originally shipped with Google Maps as standard, for example. Macs shipped with Internet Explorer as standard before Safari existed

The Google Maps situation is a great example of why this will be hard. When Apple switched to their own maps it was a huge failure resulting in a rare public apology from the company. In order to switch you have to be able to do absolutely everything that the previous solution offered without loss of quality. Given Google's competence in AI development that will be a high bar to meet.

The problem with the analogy is that users were asked to change their habits. Apple switching Siri models behind the scenes is much less problematic.

Re: Apple picks Gemini to power Siri

#55

Post by MBCook »

johnthuss wrote: post #23
Angostura wrote: post #16

Earlier quoted context omitted.

I don't see why - iOS originally shipped with Google Maps as standard, for example. Macs shipped with Internet Explorer as standard before Safari existed

The Google Maps situation is a great example of why this will be hard. When Apple switched to their own maps it was a huge failure resulting in a rare public apology from the company. In order to switch you have to be able to do absolutely everything that the previous solution offered without loss of quality. Given Google's competence in AI development that will be a high bar to meet.

They switched despite Apple Maps having poor data for a reason:

Google wanted to shove ads in it. Apple refused and to switch.

Their hand was forced by that refusal.

Re: Apple picks Gemini to power Siri

#56

Post by IOT_Apprentice »

hu3 wrote: post #8
charliebwrites wrote: post #6

Earlier quoted context omitted.

Why so? Apple explicitly acknowledged that they were using OpenAI’s GPT models before this, and now they’re quite easily switching to Google’s Gemini

I guess the question is, when are they going to use their own model? Surely research money is not the problem. Can't be lack of competence either, I think.

It appears to be lack of competence given they lied about the initial features of Apple Intelligence.

First, they touted features that no one actually built and then fired their AI figurehead “leader” who had no coherent execution plan—also, there appears to have been territorial squabbling going on, about who would build what.

How on earth did Apple Senior Management allow this to unravel? Too much focus on Services, yet ignoring their absolute failures with Siri and the bullshit that was Apple Intelligence, when AI spending is in the trillions?

Re: Apple picks Gemini to power Siri

#57

Post by Fiveplus »

drob518 wrote: post #53
Fiveplus wrote: post #42

The writing was on the wall the moment Apple stopped trying to buy their way into the server-side training game like what three years ago? Apple has the best edge inference silicon in the world (neural engine), but they have effectively zero presence in a training datacenter. They simply do not have the TPU pods or the H100 clusters to train a frontier model like Gemini 2.5 or 3.0 from scratch without burning 10 year…

Yea, I think it’s smart, too. There are multiple companies who have spent a fortune on training and are going to be increasingly interested in (desperate to?) see a return from it. Apple can choose the best of the bunch, pay less than they would have to to build it themselves, and swap to a new one if someone produces another breakthrough.

100%. It feels like Apple is perfectly happy letting the AI labs fight a race to the bottom on pricing while they keep the high-margin user relationship.

I'm curious if this officially turns the foundation model providers into the new "dumb pipes" of the tech stack?

Re: Apple picks Gemini to power Siri

#58

Post by Workaccount2 »

If nothing else, this was likely driven by Google being the most stable of the AI labs. Gemini is objectively a good model (whether it's #1 or #5 in ranking aside) so Apple can confidently deliver a good (enough) product. Also for Apple, they know their provider has ridiculously deep pockets, a good understanding and infrastructure in place for large enterprises, and a fairly diversified revenue stream.

Going with Anthropic or OpenAI, despite on the surface having that clean Apple smell and feel, carries a lot of risk Apple's part. Both companies are far underwater, liable to take risks, and liable to drown if they even fall a bit behind.

Re: Apple picks Gemini to power Siri

#60

Post by whereismyacc »

Fiveplus wrote: post #42

The writing was on the wall the moment Apple stopped trying to buy their way into the server-side training game like what three years ago? Apple has the best edge inference silicon in the world (neural engine), but they have effectively zero presence in a training datacenter. They simply do not have the TPU pods or the H100 clusters to train a frontier model like Gemini 2.5 or 3.0 from scratch without burning 10 year…

best inference silicon in the world generally or specialized to smaller models/edge?
Post reply on HN