"Assuming that they've implemented eye-tracking based UI like Vision Pro, and I don't have to shoot tiny targets to click, which is hilariously bad UI."
Assuming that the Steam Frame isn't accompanied by a complete change to the current SteamVR experience that hasn't been so much as hinted at, alas, no, SteamVR is full of tiny targets to shoot. I've only ever used the Meta Quest 3S' native UI but the smallest targets there are generally significantly bigger than the smallest targets in the SteamVR UI. On the plus side, once you activate some of those small targets you can do some cursor navigation like a conventional UI, and having that option is a breath of fresh air... but it's completely inconsistent. You experience it as a bonus when it's available because it's not even consistent enough to "miss it when it is gone", let alone for it to be a consistent navigation method.
We may get the obvious eye-tracking upgrade but the targets are still pretty small, it's going to need to be very accurate.
I don't like that aspect of emacs, but if you are a heavy-duty editor user it becomes difficult to arrange a consistent set of emacs shortcuts that aren't modal that don't conflict with anything else, because there's so many things you might want to do and so many pre-existing keyboard shortcuts that you can conflict with, not just in emacs but in your window manager. As a simple for-instance, I've got four or five keyboard shortcuts I added in the last year for dealing with the Claude windows in emacs that I've been using (the package defines a couple dozen, it's just about five I use a lot), and I didn't even try to figure out how to make them anything other than "C-c c $something" because it's hard to find somewhere they can go in any sort of pattern that makes any sense and doesn't conflict with anything. Fortunately most Unix window managers seem to leave the Windows key alone, but of course if I try to bring that to Windows it would fail miserably.
I did remap my heaviest hitters a long time ago to single strokes, though. Most notably, start macro, end macro, and replay macro all got coveted non-modal shortcuts.
"I've always hated stateful control. Always ripped out caps lock key from my boards (or later figured out remapping), same for insert mode"
You want some real fun, try the Microsoft Surface keyboard. Maybe they've fixed in a very recent version, but given how long the product line has had this problem, probably not. It has a stateful Fn key. That's right, a Fn key that works like capslock. There is no conceivable way this is a good idea. It means that if you actually want to use both "sides" of the Fn'd keys, you literally can not build muscle memory. If you hold the Fn key and press one of those keys, it'll do the "other" function, but if you just tap the Fn key, including because you had meant to press one of the keys but decided not to halfway through (which happens all the time, you just don't normally notice it because it's a completely normal thing to do that normally carries no consequences), you flip the polarity of the entire Fn key set. Now a normal press and a Fn-press do the opposite things. Until you flip it again.
This is not a "oh, as a multi-decade key user I have opinions about whether key strokes should be 68dB or 72dB" question. This is basic functioning of the keyboard. It's insane.
And, naturally, the key is "magic" and the OS can't see it. While I'm bitching, what is the deal with keyboards on new laptops needing special drivers? What the fuck is so special about your keyboard that you need drivers for it? I'll tell you what's so special about it: stuff you shouldn't be doing anyhow. My OS should be able to see and address all keys so I can remap them as needed. Your stupid special key that does your stupid special thing doesn't need to be a stupid special key. Make it a normal key and trigger your behavior in Windows, not in the hardware. Then I can use your stupid special key at least as a Meta or a Hyper or something. You don't need special drivers to have normal keys, you only need special drivers if you're doing something stupid.
So there's no fixing the Fn key on these systems because it's one of the magic keys that can't be seen by the OS at all, so it can't be remapped, it can't be turned off, it can't be locked into one state, you can't do anything. You're just stuck with a keyboard that, from your brain's point of view, randomly swaps a couple of dozen keys around.
Now I'm also on a programmable keyboard. This guy, to be precise: https://mistelkeyboard.com/products/0a26d32ac1e3b1d2af2896e0... which I split across my chair so I've got one half under each hand when it is resting comfortably. That's something you can't get a laptop keyboard to do.
I agree about the special keys. There are so many HID usages on the Consumer page for keyboards covering so many things, from application launchers to extra editing functions, and it is absurd how often almost none of them are used in favour of vendor-private usage numbers that then require vendor-specific software support.
I have a mac, and I wouldn't mind a fn-lock feature, but only from a different key combo, maybe fn+capslock. The behavior of fn (media keys or function keys) is a control panel setting, so I could probably whip up something with hammerspoon. But right now I just remap most things in my IDE so they don't require function keys.
Anyway, keyboards have needed drivers since we stopped using BIOS to read them, and fancy keyboards with macros tend to need at least a userspace daemon, but yeah this kind of thing should be as much a commodity as a VGA framebuffer is, something you just shouldn't have to fuss with. Far as I know though, USB and the *HCI zoo pretty much are that, so along with the OS's own built-in support, it should support the basic functionality of any keyboard, and provide standard means for extension. I believe the main reason any company ships a 1GB keyboard "driver" these days is the bundled shovelware and spyware.
> While I'm bitching, what is the deal with keyboards on new laptops needing special drivers?
The only laptops I've encountered with this are... Microsoft's. Somehow the Surface folks can't get enough of their critical drivers in-box with Windows to be able to re-install from stock media.
(IIRC it goes even deeper with USB drivers missing, but I gave up on Surface Laptops a few year ago)
"Also from a legal POV you can't really argue that collecting all conversations for detecting abuse patterns is "unreasonable"/"unnecessary" or similar"
It is also worth remembering that the entity that you are explaining this GDPR retention reasoning to is the government. I don't see the EU telling Anthropic or another AI company they can't do this for safety reasons... what I see is future legislation requiring them to give the EU access to these logs so they can enforce they own definitions of safety on it.
Which will work for the several weeks it takes for the other commercial providers to follow suit.
The tides are turning. AI companies are IPO'ing. They've gotten where they are by selling $5 bills for $1, to update the old VC adage. I think we can look forward to them rewriting the contracts, both literal and social, on AI going forward to capture a lot more of the value. Or, to put it in more HN-friendly terms, it may not be immediately obvious on a casual viewing, but you're looking at the beginning of the enshittification process hitting AI. The term is a bit deceptive in some sense, because it's not like anyone ever sets out with a terminal goal of making something shitty. It's downstream of trying to capture more value in the customer/vendor relationship by not giving the customer any more value than is barely necessary.
How's coding with qwen doing? The only thing that's going to stop the AI providers from extracting all the value until it's just barely worth using is the free competition.
Bedrock supports many models. Open weights models aren't far behind, maybe a year, 18 months.
Given they could have done this with data residency rules being respected and chose not to suggests all I need to know - this is for Anthropics IPO, not for user safety
>Open weights models aren't far behind, maybe a year, 18 months.
No, open weights are always a year behind +. By the time that year passes Anthropic/OpenAI/Google will have some new model that is ahead of the open models by a year.
Looking at computer security for the last 30 years, no one gives a fuck about user safety. Companies care about profits, and individuals don't care enough for strong laws.
We'll be back here in another year on HN talking about why we should give our retina sample and blood to Anthropic to use the model with a ton of people doing it. It's just the way humans are.
"This doesn't sound to me at all like how a democratic country is supposed to function."
There is a family of interesting theories, or perhaps if you prefer, simply a way of looking at history in which you don't consider the "United States" as a single governance structure that has existed back to 1776, but as a series of related, but distinct entities with distinct "social contracts" (a term laden with some philosophical baggage, here I just use it in a very general sense of what people expect from each other in various roles), and distinct theories of governance. While the later entities wrap themselves in the 1776 flag the current ruling structure is quite different from that era. From this point of view you can even go back and include the Continental Congress as the starting point of the "United States" and gain some insight into the way governance can fail as well.
I mention this because it may help free your mind up to consider how the systems really work today beyond the at-times jingoistic "Democracy!". There's a lot of flexibility in how you approach this because it's all opinion anyhow, but there is a strong case to be made that this is the "technocrat" era, in which the executive branch has been given a lot more power both by design and by the stresses of history to give more power to "experts" to deal with the radical changes the world has undergone. I think I can say something generally politically agreeable by pointing out that Congress doesn't seem to be particularly good at handling the world right now; how much worse off would it be if we still "representatives per person" numbers from 1776 and had a Congress of many thousands?
The de facto rules haven't really matched the de jure of the 1776 governance in a long time.
I am trying to keep this as neutral as possible. I have as many opinions as anyone else, but I'm just bringing up the general idea. I think it's probably good to initially just ponder based on one's own understanding of history and match it against your own ideas before you find other people handing you a theory on a platter. There's time enough for that.
> how much worse off would it be if we still "representatives per person" numbers from 1776 and had a Congress of many thousands?
Isn't that actually a major cause of the trouble? You expect Congress to deal with more and more complexities but limit the number of people (i.e. experts) who are members of it, causing them all to be generalists and moreover to have to spend more of their time campaigning rather than debating because the value of each seat is higher and correspondingly so is the effort someone will put in to take it from you and the proportion of your time you have to spend merely defending it.
Meanwhile people feel that their vote doesn't matter because a member of Congress now represents almost a million people and then ordinary people can neither affect the campaign nor get the ear of their own representative.
Suppose it actually had ten thousand members. Then they would be ordinary people. The members who are doctors would understand both medicine and medical bureaucracy. The members who are engineers would understand technology. Instead of them being lawyers whose first job is campaigning.
It doesn't matter if Congress has 5% of them who are actually engineers when 95% of the vote must inevitably come from non-engineers. Not that "engineer" is even enough. Just because I'm an "engineer" doesn't mean I can opine meaningfully on a civil engineering project any more than a civil engineer knows any more about AI than a current Congress-person.
No matter what solution you try to apply to that problem, it's not going to be solved. Try to split the legislative branch into interests so that we don't need the full branch to vote? Massive problems. Try to create a culture where Congress defers to the interest groups who know? Massive problems. A pretty solid case could be made that we've simply scaled past what a legislative approach can reasonably address.
One of the reasons I tried to write my post neutrally is that the whole situation is pretty complicated. Less neutrally, I'm fairly aware of the issues of the technocratic state and its "experts" who really aren't but get given what is basically stolen valor. On the other hand, pointing out problems is easy, trying to propose solutions is much harder. To be honest I just keep coming back to, cultures get the government they deserve. If a culture views governance as primarily about duty and obligation and honor, the structure probably doesn't matter much. If a culture doesn't view it that way, you're going to get corruption and abuse. And unfortunately, while this is a continuum it isn't balanced; to get good government requires a very positive and widespread commitment to those ideals. It doesn't take much deviation at all before you get pretty widespread corruption. It doesn't need a culture that actively values power and what it can bring you, it just takes less than massive, widespread agreement that power is more duty than privilege. If the US ever had that culture, which is debatable but possible, it really doesn't anymore.
> Suppose it actually had ten thousand members. Then they would be ordinary people.
I thought that's how it originally was supposed to be. You were supposed to be represented by a peer, not by a career politician.
And then of course, we have fun gerrymandering, where unelected officials will draw random lines to see how they can eke out more votes for their guy. Redistricting is important to do but I'm not convinced any redistricting committee in the US does it with the people's best interest at heart.
>What advantage is there in giving the unelected bureaucrats the authority to change the rules without approval, except to Congress in dodging accountability for what happens?
Why must congress do more? Most of this stuff would be state issues if not for the absurdity that is current commerce clause interpretation.
The limits of science, either experimental or observational, are not well taught. It is to the benefit of a lot of people to claim it can do a lot more than is actually possible.
Have you been in static types the whole time? It's a really, really common failure state in dynamic programming languages, the Everything Function, that started out as something simple, but then someone added a flag to make it also do this other thing, and then you need a flag to only do that other thing sometimes, and someone needed to operate on multiple things so they made a string parameter also optionally an array, and later someone allowed it to also be an object with this one method, or maybe another method if it's present because some other team implemented that before the first one and can't switch now, and before you know it it's a free-for-all of people adding flags and options and type analysis and if statements and you have a complete mess. Especially if this function is shared by many disparate teams, each of whom isn't "allowed" to break the others, though a single team can fail this way plenty fine.
You can still do this in static languages, but they do push back a bit more because you don't get the flexibility that dynamic languages offer when it comes to accepting a huge variety of different input types.
I've torn a few of these apart over the years. Never fun. Haven't tried with AI but suspect that would only be a quantitative change rather than a qualitative change. The fundamental problem with fixing these is lack of information about the exponential complexity of possible call mechanisms and the AI will have the exact same information problems I will, just faster.
Edit: One of them that I tore apart ended up being two entirely separate functions slammed together into one by historical contingency. I don't just mean that I broke the functionality down into multiple functions, that's a basic tool of how you tear these down and is nothing of note. I mean that one of the "everything functions" I tore down had two distinct calling patterns that were distinct functions that not only shouldn't have been festooned with so many options, but never should have been one function at all because they weren't even conceptually the same thing or even particularly related.
Think of it as two stages of a straight-line process, that were just jammed together because of the fact they got called at similar times, and the original writers weren't clear on the unrelated nature of the tasks and nobody was able to see it through all the obfuscation until I sat down, very deliberately, and I realized this as I was tearing it apart. I don't remember the details, I tend to remember things very conceptually and thus I have a hard time remembering the details of functions with no conceptual purity, but you can get close by thinking of the function as validating incoming parameters, and then applying the parameters to a database. And people were so confused that despite the fact this function, when tickled correctly, could do it all in one shot, sometimes, kinda, with some caveats, there were places where this function was called first to validate (with flags to shut off the application), and then to apply (with flags to shut off the validation). And to be clear, I mean, I did not realize it either even from my contact with the function over the years. It was only when I sat down with it for hours and systematically tore it down that I figured that out.
Are you citing subsidized prices, or used prices? I can only find $80 Motorolas used or locked, e.g.: https://www.amazon.com/dp/B0D323V72S and that's 4GB
12GB seems to get up into $200+, and that's still a lot of "renewed" listings.
You can find quirky little loss-leader deals here and there sometimes but I don't think you're getting 12GB of RAM for $80 on a routine basis.
Anyway it's certainly not the same phone as a flagship folding phone but for daily everyday needs more than adequate, I even was able to run multiple gig apps eg. DoorDash/Uber Eats on the 8GB model.
I will say what people consider "worth the money" varies since I bought a $1,000.00 radar detector and it's like who buys that...
I have a small shoulder bag which also holds wet wipes, a bottle of water, raincoat in case it rains etc... can't really go back to just using my pockets now.
It fits in a front pocket if you wear Dockers or equivalent, which have notoriously large pockets. It fits well in the front of a backpack. It's most akin to a slightly wide paperback book.
Assuming that the Steam Frame isn't accompanied by a complete change to the current SteamVR experience that hasn't been so much as hinted at, alas, no, SteamVR is full of tiny targets to shoot. I've only ever used the Meta Quest 3S' native UI but the smallest targets there are generally significantly bigger than the smallest targets in the SteamVR UI. On the plus side, once you activate some of those small targets you can do some cursor navigation like a conventional UI, and having that option is a breath of fresh air... but it's completely inconsistent. You experience it as a bonus when it's available because it's not even consistent enough to "miss it when it is gone", let alone for it to be a consistent navigation method.
We may get the obvious eye-tracking upgrade but the targets are still pretty small, it's going to need to be very accurate.
reply