Problem is that I think my Apple TV goes into some sort of deep idle mode where tailscale stops working. So it’s been effectively useless for me when I travel.
Check the Tailscale blog and docs for AppleTV. ISTR reading about an issue like this popping up and they had a workaround of some sort. Never happened to me.
When I read that the dude was asked to take $2b from reality labs and spend it on AI, I was shocked… that they were still spending $2b on virtual reality nonsense in 2025.
That said, from what I understand, X is working on using grok to improve the algorithm.
Meta prints money as an ad company but clearly resents being one.
VR was a ~$100B+ attempt to buy pivot, and it’s generated ~single-digit billions in revenue. The tech worked maybe, but the vibe sucked, and the problem was that people don’t want to live or work there. Also, Meta leadership personalities are toxic to a lot of people.
Now they’re doing the same thing with AI e.g., throw money at it, overpay new talent, and force an identity shift from the top. Longterm employees are still well paid, just not AI gold rush paid which is gunna create fractures.
The irony is Meta already had what most AI companies don’t in distribution, data, and monetization. AI could have been integrated into revenue products instead of treated as a second escape from ads.
You can’t typically buy your way out of your business model. Especially with a clear lack of vision. Yes, dood got lucky in a couple acquisitions, but so would you if you were throwing billions around.
Do they? It seems to me that they're just aware that social media and the internet is trendy and they need to be out there ready to control the next big thing if they want to put ads on it. Facebook has been dying for years. Instagram makes them more ad revenue per user than FB but it's not the most popular app of its class.
You realize that AI is driving huge advertising growth at Meta, right?
> Meta, the parent company of Facebook and Instagram, reported strong second-quarter 2025 earnings, driven primarily by robust advertising revenue growth. Total revenue reached US$47.52 billion, up 22% from last year, with advertising accounting for $46.56 billion, an increase of 21%, surpassing Wall Street expectations. The growth was fuelled by an 11% rise in ad impressions across Meta’s Family of Apps and a 9% increase in the average ad price. Net income climbed 36% to $18.34 billion, marking ten consecutive quarters of profit outperformance. The Family of Apps segment generated $47.15 billion in revenue and $24.97 billion in operating income, while Reality Labs posted a $4.53 billion operating loss.
> Much of this growth is credited to Meta’s AI advancements in its advertising offerings, such as smarter ad recommendations and campaign automation. Currently, over 4 million advertisers use the AI-powered Advantage+ campaigns, achieving a 22% improvement in returns. Building on this success, Meta plans to enable brands to fully create and target ads using AI by the end of 2026.
You can have access to a supercomputer for pennies, internet access for very little money, and even an m4 Mac mini for $500. You can have a raspberry pi computer for even less. And buy a monitor for a couple hundred dollars.
I feel like you’re twisting the goalposts to make your point that it has to be local compute to have access to AI. Why does it need to be local?
Update: I take it back. You can get access to AI for free.
Yep this is the first time I've disagreed with the EFF on anything civil liberties related.
My view is that there's no reason why we can't come together and come up with a rating system for websites (through HTTP headers, there are already a couple proposals, the RTA header and another W3C proposal).
Once a website just sends a header saying this is adult only content, what YOU as a user do with it is up to you. You could restrict it at the OS level (which is another thing we ALREADY have).
This would match the current system, which allows households to set their devices to block whatever they want, and the devices get metadata from the content producers.
I have to imagine porn sites don’t want an underage audience either.
So I think they would voluntarily state that their content is adult only.
This isn’t rocket science. We’ve rated content for decades. We simply can’t be obtuse about this and throw our hands up and do nothing because we can’t tell if obvious things like porn are adult only content the way we’ve done on every other medium.
I'm not a shareholder but buying equity in OpenAI seems to be a much better deal (for Disney) than making OpenAI just pay royalties, no? Seems like everyone wins, unless you think OpenAI will never amount to anything and it's all a bubble.
If you destroy your brand value in the process? This is an entertainment company... The entire value of this company is the characters and the right to make new stories with them.
Does anyone envision a scenario where OpenAI or Anthropic (or google) disappears?
I can understand the investment bubble in new infra. But even that, I’m not so sure. Right now, demand is so far outstripping supply, which is why we’re having so many conversations about energy or chips.
But yes that’s the bubble people keep talking about.
We like to pretend that this is something new. Before free porn on the internet, what did people do?
- They bought porn from stores (no anonymity)
- They rented porn from hotel rooms
- They paid for it on pay per view channels (how many kids growing up in the 90s remember watching the fuzzy scrambled Spice channel and trying to make out a boob somewhere in the garbled picture?)
- They went to movie theaters for porn! I've never actually seen this but have read about it.
NONE of those actions are anonymous. They're even documented and associated with real identities.
If you're telling me that somehow adding some kind of verifiable age gates to things like porn (and social media) will lead to authoritarianism, we don't have to theorize. We ran this experiment already. And nobody had issues with it, realistically.
We're still running this experiment. If you try to watch an R rated movie in a theater, they will likely ask you for ID if you're clearly under 17. They used to do that to me growing up.
The current state of "Kids can access all this without any protection at any time" is abnormal. It's NOT normal.
We can all have a reasonable conversation here without always bringing in the authoritarianism boogeyman.
A store that asks for ID before letting someone buy a porn mag will have an employee check with their own eyes and then return the ID. By the time the person is out the door, any record of who they were is basically gone. The record is just in the fading short-term memory of the employee. If they turned right around that instant and went back into the building, they're basically as anonymous as they were the first time. If they interrogated everyone affiliated with the store, including the owner, nobody would know the home address of the person even though it was on the ID. The anonymity is even larger if you let a few days pass. Nobody will ever remember you for having bought that magazine or for having ever entered that store for that matter. They'll likely be different employees.
With online services, the identity would be tied to your account forever. A government would have the ability to review every passing comment you've made in your entire lifetime and know exactly who you are, who your family is, where you live, where you work, what your bank accounts are, etc.
As others have pointed, this is unfortunately already true, due to big data and analytics companies hoovering up everything.
So again, what changes in any fundamental way?
This is a tech forum. There are already standards that can be used to verify age without requiring a lot of extra info. (Already used by drivers licenses at TSA checkpoints, those are all standards).
There are ways to solve this without essentially saying “there is no alternative so let’s give nefarious companies access to our kids brains all the time”
Just because it's bad, doesn't mean it's fine if it gets way worse. There's a difference between individual platforms potentially selling data with aggregators maybe finding ways to fallibly join them together, and the government mandating using a standardized identifier between all of them such that platforms can't even choose to be privacy-respecting.
You can use pseudonyms on practically all social media. That's under threat.
> There are ways to solve this without essentially saying “there is no alternative so let’s give nefarious companies access to our kids brains all the time”
It's ultimately the parents' responsibility, and the parents willingly gave their children access. It's very easy to setup parental controls. It's very easy to get your child a dumb phone. It's very easy to confiscate technology that their child isn't using responsibly. Adults should know that peer pressure isn't reason enough to give their children more freedoms than they're able to responsibly handle.
If people want to tackle this on a societal level, we should be looking into what's causing parents to be so lax in their parenting.
Right, the practical privacy outcomes are dramatically different when the digital panopticon is up and running.
The expectation of privacy from random unaffiliated humans seeing me pass on the sidewalk is very different from being stalked by a drone swarm that follows me and whatever vehicle I'm in.
As a sibling poster has said, I don't know how much on-device AI is going to matter.
I have pretty strong views on privacy, and I've generally thrown them all out in light of using AIs, because the value I get out of them is just so huge.
If Apple actually had executed on their strategy (of running models in privacy-friendly sandboxes) I feel they would've hit it out of the park. But as it stands, these are all bleeding edge technologies and you have to have your best and brightest on them. And even with seemingly infinite money, Apple doesn't seem to have delivered yet.
I hope the "yet" is important here. But judging by the various executives leaving (especially rumors of Johnny Srouji leaving), that's a huge red flag that their problem is that they're bleeding talent, and not a lack of money.
I’m much more optimistic on device-side matmul. There’s just so much of it in aggregate and the marginal cost is so low especially since you need to drive fancy graphics to the screen anyway.
Somebody will figure out how to use it—complementing Cloud-side matmul, of course—and Apple will be one of the biggest suppliers.
You don't have to abandon privacy when using an eye - use a service that accesses enterprise APIs, which have good privacy policies. I use the service from the guys who create the This day in AI podcast called smithery.ai -we are access to all of the sota models so we can flip between any model including lots of open source ones within one chat or within multiple chats and compared the same query, using various MCPs and lots of other features. If you're interested have a look at the discord to simtheory.ai (I have no connection to the service or to the creators)
That’s huge. Hope they can continue to keep such people because it isn’t just about one person. It’s all the other smart people that want to work with them.
On-device moves all compute cost (incl. electricity) to the consumer. I.e., as of 2025 that means much less battery life, a much warmer device, and much higher electricity costs. Unless the M-series can do substantially more with less this is a dead end.
That's fair for brute force (running a model on the GPU), but that's exactly where NPUs come in - they are orders of magnitude more energy-efficient for matrix operations than GPUs. Apple has been putting NPUs in every chip for years for a reason. For short, bursty tasks (answer a question, generate an image), the battery impact will be minimal. It's not 24/7 crypto mining, it's impulse load
For the occasional local LLM query, running locally probably won't make much of a dent in the battery life, smaller models like mistral-7b can run at 258 tokens/s on an iPhone 17[0].
The reason why local LLMs are unlikely to displace cloud LLMs is memory footprint, and search.
The most capable models require hundreds of GB of memory, impractical for consumer devices.
I run Qwen 3 2507 locally using llama-cpp, it's not a bad model, but I still use cloud models more, mainly due to them having good search RAG.
There are local tools for this, but they don't work as well, this might continue to improve, but I don't think it's going to get better than the API integrations with google/bing that cloud models use.
Battery isn't relevant to plugged-in devices, and in the end, electricity costs roughly the same to generate and deliver to a data center as to a home. The real cost advantage that cloud has is better amortization of hardware since you can run powerful hardware at 100% 24/7 spread across multiple people. I wouldn't bet on that continuing indefinitely, consumer hardware tends to catch up to HPC-exclusive workloads eventually.
You could have an AppleTV with 48 GB VRAM backing the local requests, but... the trend is "real computers" disappearing from homes, replaced by tablets and phones. The advantage the cloud has is Real Compute Power for the few seconds you need to process the interaction. That's not coming home any time soon.
Interestingly, some of Apple’s devices do already serve a special purpose like this in their ecosystem. The HomePod, HomePod Mini, and Apple TV act as Home Hubs for your network, which proxy WAN Apple Home requests to your IoT devices. No other Apple devices can do this.
They also already practice a concept of computational offloading with the Apple Watch and iPhone; more complicated fitness calculations, like VO2Max, rely on watch-collected data, but evidence suggests they’re calculated on the phone (new VO2Max algorithms are implemented when you update iOS, not watchOS)
So yeah; I can imagine a future where Apple devices could offload substantial AI requests to other devices on your Apple account, to optimize for both power consumption (plugged in versus battery) and speed (if you have a more powerful Mac versus your iPhone). There’s good precedent in the Apple ecosystem for this. Then, of course, the highest tier of requests are processed in their private cloud.
If the cloud AI is ad or VC-supported, sure, but that doesn't seem like a sustainable way to provide good user experience.
And don't worry, I'm sure some enterprising electricity company is working out how to give you free electricity in exchange for beaming more ads into your home.
What is the possible benefit of on device processing?
I envy your very simple, sedentary life where you are never outside of a high-speed wifi bubble.
Look at almost every Apple ad: It's people climbing rocks, surfing, skiing, enjoying majestic vistas, and all those things that very often come with reduced or zero connectivity.
Apple runs all the heavy compute stuff overnight when your device is plugged in. The cost of the electricity is effectively nothing. And there is no impact on your battery life or device performance.
My favorite was Is it time to rewrite sudo in Zig. As you may remember Ubuntu had a recent vulnerability caused by their sudo being rewritten in Rust. No idea if the AI knew that though it feels like a clever joke.
And which number gets you there depends on your lifestyle.
And taking a job without any consideration of pay.
reply