Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Google Gemini Partnership With Apple Will Go Beyond Siri Revamp
- Thread starter MacRumors
- Start date
- Sort by reaction score
The answer is a lot simpler. Just put in the Apple User Agreement something like, "If your data is leaked by Apple or its subcontractors Apple will compensate the User a minimum of $100,000".
This makes it an economic issue that no one at Apple would want to cross.
Use the word privacy often enough and the public will believe it. Especially those living in a TLDR world.
For the critical thinkers who care about privacy and read the terms and conditions, it's an eye opening experience leaving one well informed.
Once educated, the reader realizes that the difference between Apple and Google is like splitting hairs.
For those who embrace and use AI often, a new Google Pixel with Gemini will put a Google Gemini powered iPhone to shame.
I literally did — I laid it out for you — a requirement of public transparency, the usage of Private Cloud Compute, and the additional incentive third-parties have to prove someone who vocally makes claims of privacy to not only be caught lying, but actually has a reward set up for anyone who finds issues.
Again — you are choosing to ignore information to confirm your bias. They have indicated what tech they are using — and those who actually understand it have had over a year to poke it with a stick and had no vulnerabilities to date, or lapses in transparency. The absence of evidence is not evidence, my friend — if you want to prove something is not as claimed, the onus is on you to prove it, and a double negative isn’t proof of anything other than your opinion based on feelings and not facts. I’ve provided facts — there are entire white papers walking you through the process with industry standards.
As for the example I offered, it actually strengthens my argument. The process of recordings for training was standard practice in the industry — yet only Apple being called out made the news (because Microsoft, Google and Samsung were also caught at the same time, but didn’t garner as much interest) proves my point that if it is Apple doing it — it’s newsworthy — so Apple has to take extra care in something they have leveraged so heavily... their claim of privacy and security. They have to be twice as cautious as their name would make twice as large of a news splash.
It is incredibly easy to be contrarian when you don’t have to offer anything other than feelings to support an opinion.
You should maybe ready the Usage agreement, because yes — it is in there. Even if you Opt in, it still has heavy anonymization.
Last edited:
Read the information — Google won’t be holding any data, hosting any of the software, connecting to any of it’s services, nor involved in transmission of data. It is exactly like how it works with anonymous ChatGPT if not logged into an individual account — an extension that runs locally — and if anything needs to be sent out — it only goes to Apple's Private Cloud Compute... Not any Google servers. Literally all they are using is the machine language model, nothing else. It isn’t just sending the request to Google’s Gemini implementation.
I understand your arguments but they rely on one trusting a company ... any company. The companies we are talking about are huge, and not all their employees, even (especially?) at the senior level, are necessarily honest. There's a lot of intra-company politics that might tempt some to take shortcuts to "success" within the company.
In Apple's case, there are numerous examples of successful lawsuits against them in which they stole ideas, or failed to pay royalties on inventions or previously-licensed software. I have experience of a particular case, and ultimately this was due to a small group of employees thinking they could save a little money for Apple by stopping agreed licence payments for SW used in the then-current iPhone models. It was not official Apple policy to be dishonest, but it happened anyway.
That's the problem ... can we trust what companies say about privacy?
Shanghaichica
macrumors G5
With respect, you are conflating issues. I agree — people shouldn’t just trust companies at their word — but that’s not what is happening here. From a privacy standpoint, apple has provided third party transparency for Private Cloud Compute — so they made a "framework for allowing independent security experts and researchers to verify that user data remains private during AI processing". This initiative moves beyond traditional "trust us" cloud models by enabling external validation of Apple’s security promises, such as stateless processing and data encryption. It can be independently verified. Apple publishes cryptographic, tamper-proof logs of all code running on PCC nodes. They make the binary software images (operating systems, applications, and executables) used in production available for independent inspection. Apple provides a VRE that simulates the PCC environment, allowing researchers to analyze the software, identify vulnerabilities, and verify functionality. Also, to incentivize inspection, Apple offers up to $1 million for identifying significant security or privacy breaches in the PCC system. They go beyond the industry standard to provide transparency in this area — they shouldn’t get a pass on other things because they are doing things right here but with the same token they shouldn’t be punished here for things they have done in the past in other avenues.
You have valid points (and I share them) about these other situations — and not just about apple, but all companies — but this is not that. We have to put our biases aside and judge situations in isolation. On this, they are approaching privacy correctly, and better than other companies by a large margin.
I defer to your understanding of this topic, which is far deeper than mine.
But I'm still uncomfortable about the possibility of internal actions inside Apple, where employees do not follow company policy (at least, stated company policy).
I get it — and by no means am I trying to convince you (or anyone) to blindly have faith that a company would just “do the right thing”. All I suggest is to take matters in isolation. In the very specific instance, Apple has done an admirable step in transparency on this specific topic — and if we use critical thinking on this very important variable in how Apple’s AI process will work moving forward, the transparency they are offering is unlike what anyone in the industry is putting into process. They are opening themselves up to a level of scrutiny (due to the way the modulated their AI system) that has a much higher potential on negative feedback than positive — meaning the bar is much higher due to the ability of checking their work — so this in no way helps them if they are looking to do something shady, they will be found out.
If they didn’t believe in their ability to both maintain privacy and to eliminate transmission of personal identifiers, having this system in place places an unnecessary burden on themselves when no one in the industry does the same thing.