Not that I’ve seen, but I know some people who somehow missed the video, and he doesn’t link to it on the website so:

https://youtu.be/w70Xc9CStoE

I bet if such a law existed in less than a month all those AI developers would very quickly abandon the “oh no you see it’s impossible to completely avoid hallucinations for you see the math is just too complex tee hee” and would actually fix this.

Nah, this problem is actually too hard to solve with LLMs. They don’t have any structure or understanding of what they’re saying so there’s no way to write better guardrails… Unless you build some other system that tries to make sense of what the LLM says, but that approaches the difficulty of just building an intelligent agent in the first place.

So no, if this law came into effect, people would just stop using AI. It’s too cavalier. And imo, they probably should stop for cases like this unless it has direct human oversight of everything coming out of it. Which also, probably just wouldn’t happen.

My experience is dated, but figured I’d share it in case no one else has any input.

I owned a few Motorola Android phones before and after the Google involvement. I think my most recent purchase was 2015.

At that time, they were extremely “pure android” with very few additions beyond the stock experience. The things they added were way ahead of their time - I still think those devices had the best “always on” display implementation to this day, and they did it way before it became a norm.

Their software and update support was rivaling Google at the time, and most other manufacturers were still in the days of 2 years of updates if you’re lucky.

They just stopped making phones it seemed like. I ended up moving towards Pixels over the years, but Moto is the one company that would tempt me to switch back. That or maybe HTC but they’re dead.

Hope you get a more recent answer - I didn’t even realize they were still making phones to be honest.

Yes, because when you run systems like that, you use the AI, and you have the people as a fallback for when the AI fails.

It was primarily watched by people in India because the AI was failing the vast majority of the time.

So yeah, the state of the art AI is… Failing at its job 70% of the time. Instead of the hoped goal of 5%.

Yeah that’s totally valid. Agreed.

But I also wouldn’t really trust third party parts for the appliance itself. I think once you do, that immediately becomes a possible problem. If it was in my house, I’d only buy from the manufacturer for something like that.

But on the other hand, Idk that it’s necessarily wrong to legislate forcing these companies to allow it. I generally believe consumers should have the option on their own, but some things are too dangerous. I’d pretty much be against medical devices but HVAC is a little more uncertain to me.

I mean, I don’t want the thing supplying the air I’m breathing to accidentally not burn all the gas and lead to carbon monoxide poisoning etc… Things like the ductwork and shit, for sure, but not like, a burner.

I could see an argument about medical devices, HVAC, and vehicles… But I don’t think I’d agree with them. Except maybe medical.

Consoles and toothbrushes though? What the fuck?

There’s isn’t enough physical space for three sensors on a smaller phone especially if it’s the size of the iPhone mini

I wouldn’t go as far as to claim that “more cameras” is the complaints being made here. Sure, telephotos make sense as things that take up more space. But most people are using them for like 1 in 50 shots or something. I have an extremely hard time believing that someone would genuinely notice the difference unless they’re an extreme case or they’ve been told the other ones are better. Within reasonable effective focal lengths, these are pretty negligible in the sizes we’re talking about.

If Apple couldn’t make a smaller phone sell particularly well, I doubt anyone else could.

I hard disagree with this. Apple is literally the worst company to try to make this shit work. Apple’s core selling point is the status symbol of it all. People trying to show off having the flashiest phone are not going to buy a product being touted as a half baked smaller and cheaper version of something else. Their entire marketing was about it being mini. Apple customers are not the core audience for something like this, and Apple marketed it as exactly what people disliked about small phones.

around or less than 5% of total iPhone 12 and 13 sales

I find it more surprising that this was below expectations than I do that only 5% of people bought a smaller phone. I doubt much more than 1 in 20 people really is after a smaller phone. I’m sure they exist, but based on the people I know and the number of people I’ve heard interested in smaller phones, I’d estimate it more like 1 in 20 to 1 in 40. It’s not for most people by any means. But 1 in 20 is still a decent number of people.

They have to understand that the cameras on the biggest flagships occupy a lot of space and it isn’t feasible to bring it to a smaller form factor.

Not… Really… Sure it makes some difference, but the much more constraining factor is the money. Cameras arent that big, but they’re one of the priciest pieces of hardware in the device.

The problem is more that they keep trying to sell small phones at cheaper price points. So they end up with much worse screens, socs, and cameras so they perform like shit. People don’t want a small phone because they don’t care about their phone. People want small phones because the standard size is fucking huge. They need to make a high-ish tier small phone instead of low tier small phone that performs like the 50 Walmart shit.

I know it’s not everyones cup of tea but…

I ended up buying a Pixel Fold for the Fold aspect. There’s an added benefit that it’s really short when folded so I can reach the entire screen one handed. It’s heavier and thicker than a slate phone, but I actually enjoy the folded experience way better than other phones.

This seems like a detour around right to repair.

That’s because it is. That’s all Apple does. Every time they get brought to court around shit, like the app store stuff in EU that just happened, they make it intentionally as difficult as could possibly be while still technically following the request. It’s malicious compliance at every step of the way even when they get caught. They’re so anti-consumer it’s not even funny.

Yeah, it’s in the name is “security”. As if a third party camera or back cover is going to break into the OS, harvest super important user data, and then somehow find some way to transmit it back to headquarters.

You know, or they just want to make money off of selling parts at 200% profit instead of Apples 500%.

The idea that this is somehow a security risk is a giant steamy pile of bullshit to keep people buying their garbage.

Yes, but Chromebooks are far from “newcomers” these days. They’ve been out a while. Many people who grew up using them in schools are now making their own purchasing decisions, etc.

Like he said as the second sentence of his comment…

You’ve never worked in enterprise then.

These solutions are skipping the majority of the core problems he mentioned. And even the problem you’re trying to solve here isn’t even fully solved by this solution. You’re taking a narrow sliver of one point in his argument and arguing about that and just tossing out the rest. Even if we accepted your proposal, Linux still isn’t enough of an answer here.

While I don’t really disagree, look at the market share of Chromebooks. If “most people” only needed internet access, “most people” would be on Chromebooks by now. It’s not like they’re unknown anymore.

Yeah, and it’s likely way less costly to the company to just buy a new win 11 computer than it is to pay an employee to train on new software. Not to mention the cost of paying someone to find someone to do a Linux conversion, paying the person doing the conversion, and the loss of productivity as the person learns. Not to mention the cost of changing IT infrastructure, hiring new IT people to manage those machines, etc.

There’s a reason companies don’t just switch at the drop of the hat. There’s too much commitment and institutional knowledge already and moving is not a simple change.

I can’t agree with this more. People like to sell Linux as a magic bullet, but it does not and will not everything everyone needs without maintenance and people really like to hand wave or downplay that need.

Sure, you could find a solution for what they’re using now. What happens when they need something else and they’re so tech illiterate that they don’t even know what you did to their machine? They wouldn’t even know how to install new software, and if they did, they wouldn’t know they need to click the Linux version, etc. It’s not always about feasibility and available options, it’s often about the fact that people just won’t fucking know what to do. Even if you assume there are enough options available, they won’t know how to do so.

And every step Microsoft takes to shoot themselves in the foot, and every step Linux takes to make this easier, everyone comes screaming about how much this could change things.

But until Linux has a HUGE market share - like in the 30-70 percent range - developers are not going to take it seriously and alleviate this process. Even with how well MacOS does, this is not even a solved problem entirely there - there are still hang ups and still software that doesn’t get released for mac. Linux would have to pass where Apple is today for this to become remotely accessible to an every day person.

And even THEN there’s the question of different Linux distros.

I really do not understand how server anti cheat is not way easier.

In a clean slate, it is. It’s also way more effective (except for things like wall hacks, aim bots, recoil suppressors, etc, but most of those things are only really important and popular in competitive FPS). It’s also much simpler to understand and to leave no “holes” behind. It also lives in the developers domain so it can’t be “compromised” or circumvented.

The thing is that client side “anti cheat” can be commoditized. Every game with server authority/anti cheat needs specific server software to run their game logic. Client anti cheat is basically “look at everything else running on the system and see if any of it seems suspicious”. As such, there’s not really anything “game specific” to these - they basically are just a watch dog looking for bad actors - so as such, one company can come along, make one, and sell it to other devs.

This being “off the shelf” and not something the dev team has to think about besides a price tag means that management is just going to buy a third party solution and check off the “anti cheat” box on their task list.

I feel like devs are caught up on realtime anti cheat and not willing to do anything asynchronous.

First, this is a management problem and not the devs. Any dev worth their salt knows this isn’t really a good solution.

But I’d say the more relevant and prominent thing here is that game companies just don’t want to have to run servers anymore. It’s a cost, requires dev time, and requires maintenance, and they don’t want to do that. If these games had servers running the game world like games used to, they’d inherently have their own “anti cheat” built in for free that wouldn’t necessarily catch everything but would do a better job than some of these. And it could be enhanced to cover more bases.

But studios don’t want to do this anymore. It’s easier to make the game p2p and slap an off the shelf anti cheat and call it a day.

Some games still require matchmaking servers etc, but the overhead there is way lower.

Or they really like paying licensing fees for client-side anticheat.

Not that I agree with the decision, but it is definitely cheaper and faster than the alternative. But picking something like nprotect totally fucking baffles me. There are better options.

I just don’t understand how any competent software engineer or systems admin or architect trusts the client so fervently.

In some ways, same. Every project I’ve been on that has gotten anywhere near client side trust I’ve fought adamantly about avoiding it. I’ve won most arguments on it, but there are some places where they just utterly refuse.

But then there are things like New World… I don’t know how the fuck that shit released like it did. The number of things trusted to the client were absolutely baffling. I expected Amazon’s first foray into gaming to be a fucking joke, but I was totally appalled at how bad it turned out. They even touted hiring ex blizzard talent to get my hopes up first.

It does

https://steamdb.info/tech/AntiCheat/nProtect_GameGuard/

https://www.resetera.com/threads/helldivers-2-incorporates-nprotect-gameguard-on-pc-a-kernel-level-rootkit-anti-cheat-technical-director-responds-to-the-controversy-on-reddit.808530/

I don’t know if this makes me “a redditor” somehow or what, but…

As a dev, I am deeply troubled by the gaming industry so calmly walking into kernel anti cheats. It’s insane and being tossed around like it’s nothing.

Helldivers especially, since they picked one of the sketchiest ones and it’s a game that entirely doesn’t need it.

I have no idea if Reddit has suddenly picked up on this, but I’ve been pissed since at least Valorants release, but have seen more YT videos talking about it recently.