My view of AI (2)

[BRBFBI]

Didn't want to start a new thread for this, but have you guys seen the Apple Intelligence ads from a couple months ago?

They're all bad, most of them with protagonists who are explicitly lazy and unproductive using Apple Intelligence to, at least on a surface level, make up for their shortcomings. However, this one in particular disturbed me fairly deeply.

I already have a hard time with gifts. I'm somewhat of a minimalist and, as an adult, anything I actually want I can just buy myself. The barrier to getting a new object isn't whether I can afford it, but whether it's worth its spot in my home, so when someone gives me a generic gift I usually just feel bad because except on rare occasions it's going to end up at the thrift store. My favorite gifts are ones that may be cheap or goofy but show that the giver really knows me or put some effort into making it (and I can toss later, but with a smile).

This ad is the opposite of that. She uses Apple Intelligence to essentially manipulate her husband into feeling emotional because he thinks she put this slideshow together for him. Obviously this requires a bit of suspension of disbelief since iPhoto's been making trashy slideshows nobody asked for for years now, but that's not the point. She manipulated her husband emotionally and the ad tells us this is a good thing.

Not only am I a minimalist but I'm also somewhat of a romantic, and the idea that the consumer would want to offload aspects of their relationships and simply "go through the motions" if they could get away with it is existentially horrifying. I feel genuinely bad for the husband. In the world presented in the ad a part of you would always have to wonder if your loved ones actually spent effort on you, or 30 seconds prompting AI to do it. I truly cannot believe that this is a real ad. It's like a cautionary tale with the moral punchline removed.

I guess this does fit pretty well in this thread:

"It's sorta tragic
It does it all for you
There's nothing left to do

I do take consolation in the fact that the video has 8K likes and 28K dislikes (return youtube dislike firefox extension).

[chuft]

I avoid both ads and AI like the plague, so I didn't see it, but I agree with you on all points.

[chuft]

AI haters build tarpits to trap and trick AI scrapers that ignore robots.txt

[chuft]

"The Google owner, Alphabet, has dropped its promise not to use artificial intelligence for purposes such as developing weapons and surveillance tools."


https://www.theguardian.com/technolo...ai-for-weapons

[possessor]

why would they have to promise not to do that in the first place

[chuft]

Because it's evil?

You can read their original promise here

https://web.archive.org/web/20250130...ty/principles/

[BRBFBI]

I heard that on the radio on my commute this morning. Those sound like promises from a more naive time. It's inevitable AI will be used for weapons and surveillance and if it's not Google making money off it somebody else will. We need regulation, not empty promises from for-profit businesses.

[boredjedi]

Remember Google's Logo "Google Don't Be Evil"? That didn't last either.

[chuft]

I don't think ethics are naive. And that was first posted in 2023, not very long ago at all. I don't want to end up in some creepy surveillance state like China has so Google can make even more money.

[BRBFBI]

I don't think ethics are naive. And that was first posted in 2023, not very long ago at all. I don't want to end up in some creepy surveillance state like China has so Google can make even more money.

Neither do I want to live in a surveillance state. I don't think ethics are naive, but Google was naive to think they (and a small number of other US companies) had a technological moat around AI large enough that they could shape how AI would be used. And I think people who had faith in Google's promises were naive to the way that a publicly traded corporation's singular goal is to return value to their shareholders. Any code of ethics that gets in the way of that is going to quickly go away.

[chuft]

I don't think that's true. Most if not all companies of any size have codes of ethics. I think a perfect comparison to Google is Apple.


Our Values


“We believe that business, at its best, serves the public good, empowers people around the world, and binds us together as never before.”


-Apple CEO Tim Cook


Apple are the good guys. Anybody in the privacy field will tell you Apple is by far the best company if you want to keep your information private. The encryption keys for things like passwords and text messages are kept on the user's device. iMessages have always had end to end encryption. Apple does not have access to them. Google on the other hand is in the business of selling your information. They have been caught selling Gmail user content to outside companies, in addition to facilitating cross-website tracking of users.

There are many ethical companies out there. Google is not one of them. They used to be better - they left China rather than let Chinese censorship and hacking of activists' Gmail accounts continue - but they have gone down a dark hole since then.

If companies were just focused on returning value to shareholders, they wouldn't pay their executives so much....

A good code of ethics among other things keeps companies from doing things that land them in the news with bad publicity, which is bad for the company's reputation and thus hurts shareholders. It also lets you attract better quality employees.

[BRBFBI]

A good code of ethics among other things keeps companies from doing things that land them in the news with bad publicity, which is bad for the company's reputation and thus hurts shareholders. It also lets you attract better quality employees.

I think that's true, and I also think it requires a strong leadership team to take the long view. I've been surprised how shortsighted some corporate decisions can be (e.g. Google making their search results worse to increase ad views [no matter how hard I search I cannot seem to find your post on that right now]). I think in the business world with the revolving door of CEOs there's an incentive to do whatever will boost quarterly profits and give hard numbers to show the board of directors. Taking the long view requires conviction that I don't think a lot of corporate structures are designed to accept. I suppose that was that was the point of your above post: that Google is hurting their reputation to make some money on AI for weapons and surveillance tools. My point is that I'm not hedging my bets on a corporation doing the right thing.

[BRBFBI]

On the topic of Apple:

I have an iPhone. As someone who likes having control over my files they've always rubbed me the wrong way (I hate how hard it is to move photos on and off of them, and they don't let you expand your storage with a microSD card like Andriod does). If they didn't have a good reputation for security that would have been all the excuse I'd need to leave them.

However, I've heard that Apple Intelligence (starting with iPhone 16) potentially circumvents encryption.

To explain this, let's go back to Apple's plan to scan iCloud photos for CSAM. To do this, Apple would use a hashing algorithm included in your device's OS. When a user uploads an image to iCloud the device first hashes the photo and compares it to an on-device database of known CSAM hashes. If there is a match it sends this information along with the photo to iCloud. "Apple says the system does not work for users who have iCloud Photos disabled, though it’s not totally clear if that scanning is only performed on images uploaded to iCloud Photos, or all images are scanned and compared but the results of the scan (a hash match or not) are only sent along with the photo when it’s uploaded to iCloud Photos (MacWorld).

To be clear, this isn't a "smart" (AI) system, it's a simple hashing program, but it circumvents encryption by scanning photos on a user's device then sending the results of the scan to Apple. This type of system is known as Client Side Scanning. While the intention was to detect CSAM, there is nothing technology-wise stopping Apple from adding hashes for known images of, say, Tienanmen Square or LGBT content in a software update.

Apple recognized the potential for abuse of this system and, in the end, chose not to implement it. According to Erik Neuenschwander, Apple's director of user privacy and child safety, “Scanning every user’s privately stored iCloud data would create new threat vectors for data thieves to find and exploit. It would also inject the potential for a slippery slope of unintended consequences. Scanning for one type of content, for instance, opens the door for bulk surveillance and could create a desire to search other encrypted messaging systems across content types.”

Which brings us to Apple Intelligence. Unlike a hashing algorithm which has a very specific function, Apple Intelligence processes everything on your device all the time. Some claims about Apple Intelligence:

-Writing Tools: AI will be able to rewrite and proofread your text anywhere you write it, even third-party apps. This will be especially useful when sending an email or writing an article or social media post.

-Onscreen awareness: Now Siri will know what’s on your screen and can assist accordingly. For example, you might say, “Add this address to their contact card.”

-Personalized assistance: Siri can perform new actions with a deep awareness of your personal context. For example, you can say, “Play that song my wife sent me the other day.”

-Summaries of content: Whether in a transcript from a meeting or when viewing an email, AI will summarize content for you. (9to5mac)

This is essentially the hashing algorithm but for everything that happens on your device. If you can see it or hear it, so can Apple Intelligence. While iMessage might still be encrypted and apple can still (correctly) claim they can't decrypt your messages, it is circumvented if Apple Intelligence can read them directly from your device. Theoretically a government could demand that Apple Intelligence flag users who discuss, view, engage in, or display tendencies toward anything deemed illegal. In the context of the old CSAM scanning system "Apple claimed that it would never have allowed this [government overreach], but the promise was predicated on Apple having the legal freedom to refuse, which would simply not be the case. In China, for example, Apple has been legally required to remove VPN, news, and other apps, and to store the iCloud data of Chinese citizens on a server owned by a government-controlled company." (9to5mac)

Now, as you've pointed out, Apple has a great track record of protecting user privacy, but as I understand it Microsoft Copilot+ PCs will basically be the same thing and I don't trust them nearly as much, which is why I've been using Linux for years. Regardless of who you trust, Client Side Scanning is a disturbing development for digital privacy. I suspect you already knew a lot of this, but it was helpful for me to collect my thoughts by writing them out. I'm attaching a research paper on Client Side Scanning in case you want to do some followup reading.

[chuft]

I think AI is the devil and I would disable it the instant I got a new iPhone (I have a 15 which thankfully can't use it).

I didn't know that about how it would work, but I never looked into it because I have zero interest in using AI of any kind, especially on my personal devices, from Apple or anybody else. By its nature it is evil in every way, built on theft and spying.

I am using Windows 10 at the moment, but am seriously considering moving to Linux rather than Windows 11. My summer project may be building a Linux PC.



It is quite easy to download photos from an iPhone to Windows, I don't know about Linux. It used to be easy to get other documents on and off it like PDFs, then they took that away as part of the enshittification of iTunes to try to get you to use iCloud and Apple Music. I can access my photos via Windows Explorer, or I can use a utility that comes with Windows to download photos, which shows a thumbnail of each one and lets you select which to download. It used to remember which ones you had already downloaded but it seems to have lost that feature or it doesn't work reliably.

But in general Apple's philosophy in the past decade has been "everybody should use wireless and the cloud for everything" - they dislike putting ports on Macs for example and stopped including CD/DVD drives long before anybody else - so they just expect that you will use iCloud, so you can easily use it to transfer photos between devices that way. Microsoft is the same with OneDrive. All companies have recognized the profitability of subscription services and try to push them.

In fairness, all my younger coworkers look at me like I'm from another planet when I mention I rip my old CDs and load MP3's on my iPhone. They all stream and listen to whatever the algorithm feeds them, or pay for the (temporary) privilege of getting to pick what they want to listen to.

[chuft]

Regarding why Google Search now sucks, I had put that in a private message, sorry if I caused confusion by acting as if it was public.

Some articles on it:

The Man Who Killed Google Search

The Rot Economy


This week I was searching for medications which caused a particular side effect. I tried 15 times on Google and went through 150 results or whatever it was. Then I tried Bing and found exactly what I was looking for in the very first result on the first page. A reminder of how good Google used to be, but now isn't.