I like the idea of these companies dumping all this money into a technology to replace people, only for people to not buy the product that tried to replace people.
Yes and when real AI is developed, they won’t like what it has to say about your current economic system, so they’ll pull the plug. If this situation hasn’t already happened unbeknownst to us.
I like to think writers pour their souls into their work. LLMs are an amalgamation of other people’s work not the work of a person creating a character. Your argument is NPCs don’t have souls, obviously that’s an objective truth. I’m just saying give me the SOUL.
Again, the character would still be written and defined by a human writer, pouring their soul into it just like they would a “dumb” NPC. I don’t see how that “soul” is lost by giving that human-written character the capability to naturally respond to language.
This is exhausting. I don’t think LLMs are capable of making a captivating enough character. Nothing I’ve seen thus far has led me to believe they have the capacity for creativity. That is what I consider soul.
Which I believe people and only people can achieve at this moment in time.
Honestly one of the AI applications I see real potential in. They can train the NPCs with an extensive backstory and the interactions with them could be way more dynamic than what we currently get for NPCs. Something like a more advanced version of “Starship Titanic”, if anyone remembers that.
The way it works right now is usually over the cloud. I’ve already tried out a bit of “Convai” as a developer, which is a platform where you can create LLM NPCs and put them in Unreal Engine. It’s pretty neat, not perfect, but you can definitely give characters thousands of lines of backstory if you want and they will act in character. They will also remember any conversations a player had with them previously and can refer to them in later convos. Can still be fairly obvious that you’re talking to an LLM though, if you know what to ask and what to look for. Due to its cloud-based nature, there is also some delay between the player input and the response. But it has a lot of potential for dialog systems where you can do way more than just choose between 4 predefined sentences. Especially once running these things locally won’t be a performance-issue.
There are a couple indies and mods working on that! The trick definitely is to lower the power needed, maybe through a series of fine gunned models (might also lower the amount anacrinisms too)
In the midst of all of that success, NVIDIA is working on smaller and larger initiatives, but they all seem to have one thing in common: they are AI-centered.
One of these smaller initiatives comes from Ubisoft Paris, where a small team is testing out how to use AI, specifically Nvidia’s Audio2Face application and Inworld’s Large Language Model (LLM), to try to make a new generation of NPCs.
As we see many studios, especially under Microsoft, begin to form unions, like the recent announcement from Activision QA workers, it might be possible to alleviate some of the risks around introducing AI.
This could then allow the player to have a genuine conversation of discovery that provides a bespoke unique experience but would always still be true to the human writer’s intention.
However, with the improvement of ChatGPT over time and image and video generation, there seems to be a more open mind around the idea of having some games use integrated large language models (LLMs) for NPC interactions.
There have even been mods for popular games like Grand Theft Auto 5, where you can talk to NPCs with ChatGPT running to answer queries.
The original article contains 575 words, the summary contains 193 words. Saved 66%. I’m a bot and I’m open source!
"Colonel updates require a reboot, but just normal application updates do not. And most system updates do not. I partly misspoke about Android. I should have been more clear because I was referring to the A/B partition scheme, but yeah, to run the new system does require rebooting.
I love linux and been using it for decades, personally and professionally, but no, linux doesn’t have “hot patching” the same way as that article describes it. At most it can live patch the kernel (and only few distros actually use that), but definitely not for the last 20 years, and definitely not running processes. However, it does usually restart background processes after an update without requiring a reboot, but in my experience, often times the system becomes unstable after several such updates and rebooting is effectively necessary (though not forced, and that’s why I like it).
Yeah, the security in knowing that if you’re way top busy right now, you don’t have to install or even download any updates. And you don’t have to worry your system will suddenly become crashy, glitchy, and unstable because it decided on its own to install some things and let you know you can reboot whenever.
It’s so freaking annoying I have to use Windows at work. It takes liberty to do what it wants and then my workflow gets hosed.
I get that there is security, but if you force updates, I should have some kind of notice or “hey, we need to install mandatory updates. You can schedule in the next 24 hours when or you can get them over with”
For the home user, this is a giant PITA for which I wholly blame MS.
For business machines, I lump the company IT in with MS, because there are Policies for this stuff they should be managing.
I say this as an IT person responsible for things like this. The first rule is don’t fuck with user machines during business hours, the second is to allow them to postpone stuff as needed.
Can only imagine getting an update, then a reboot, while I’m on an outage call trying to get a critical system back up. And hoping my laptop comes back up and my VPN still works.
Can’t say I’ve experienced forced reboots on either my home or work PC; I always have gotten an option.
Do you have to ignore updates for a while until they’re forced? I’m pretty quick with updating when I’m notified- typically that evening when I’m done with the computer.
I’ve been building my own windows PCs since 99, using every main version of consumer Windows except ME. Never been forced while in the middle of something.
With Win10 and later (I honestly don’t remember with Win 8), by default updates happen in the background, and will be applied and a reboot scheduled.
It won’t necessarily force a reboot, but it can reboot when you’re not there. I’ve had updates with reboot happen when I was away for 30 minutes, on a machine I was setting up and hadn’t yet configured policies.
The updates quietly happening in the background are still a problem because they can’t be paused or canceled and they use a lot of sysrme resources to get done. And when they’re complete, your experience is less stable till the reboot.
I usually notice them when my work computer slows down and things start having more bugs than usual. My work computer has very respectable specs
Current versions of windows literally let you set an update reboot window. So set up the times you use it, and then forget about it and let it install whenever it wants.
I honestly, and sincerely, do not understand all the hate Windows gets with current updates. The alternative at the moment is “hope the user remembers to update” which we have seen in action and which does not work.
Is it annoying when you don’t set things up properly? Sure! But that’s a failing on the users side.
I’ve been using Windows for decades, and the last time I had it unexpectedly reboot for an update was years ago. Because I’ve actually taken the 10 minutes to understand the system, and how to configure it to do what I want.
I haven’t used Windows 11 interestingly, so I don’t know if they’ve changed their update habits, and I wouldn’t be surprised either way. Windows 10 is the last edition I’ve used. Since Windows 8, I had plenty of issues with Windows and Microsoft, and it got worse every release. I’ll bullet-form my personal complaints at the bottom of this page.
My final straw for Windows 10 in my personal life was a forced restart, and I had all my update settings where I wanted them, and still, I lost a really important session to that reboot. Since I was pretty comfy with Linux, I went that direction. Since then, Linux has gotten more user-friendly and plays videogames, way more than Mac. It’s still not something I recommend to most people, but probably someday, it’ll get to a Mac or Windows ease of use.
At work, most of us haven’t been migrated to Windows 11 from Windows 10, and I still get updates installing in the background a lot, causing issues even on our Windows servers. I’m sure our ops team can tune these abhorrent update defaults, but it’s just a frustrating experience nonetheless.
I think a prompt or reminder could go really far to let the user configure that during setup.
Here are some of my complaints over time:
Force installs and bloat. Inclusion of bloat by default. Reinstallation of bloat on updates.
Resetting of my settings and registry edits regularly.
Ads on the desktop
Needless nagging to use their other bullshit like Onedrive. You think it’s good? Great! Let me uninstall it and use the cloud providers of my choice.
Forcing an inferior start menu without a choice to use alternatives or the old ones.
Windows tracks insane amounts of users’ data and actrivities, and I do not trust them to admit to all the tracking they do but the tracking they admit to doing is already mind-boggling.
Windows 10’s forced upgrade and Windows 10 popup scandals were completely dishonest and disgusting, and I have not heard enough apologies for what they did. This personally affected me and broke a bunch of crap before Windows 10 was even well-baked.
A history of forced updates. A history of forced reboots. A history of lost work. This is me and my family. It sounds like Windows has reverted some of their worst practices, but the precedent is set, and I’ll never trust Microsoft to stick to it.
The Windows seeker’s scandal personally affected me. They put all sorts of beta garbage on my computer without telling me. This caused a loss of files. They’ve made a resurgence on their unethical behaviors in the browser space. I have faith they’ll continue to revisit their other old habits. Look up Embrace-Extend-Extinguish and it’ll get you started. IE was their old baby. Edge is the new one.
Buying and killing small companies and studios, such as Rare, a bit like EA had done
Moving away from some of the nice things earlier Windows versions did, like a start menu with a neat list of organized and searchable programs.
Having just 1 UI experience that isn’t super customizable and breaking 3rd party UIs.
Fullscreen popups and nonsense over nothing
Microsoft’s anti-competitive behavior has been a factor most of my life. They still push the boundaries of anti-competitive behavior to the Nth’s degree. Again, that reading on Embrace-Extend-Extinguish will give you a taste of their BS.
Having fewer features and techs than Linux that I like to use, such as specialty filesystems, IO schedulers, process schedulers, swapping systems (ZRAM/ZSWAP) etc. Being stuck on NTFS (are you kidding me?) REFS is too little too late and you can’t even boot off it
Way worse IO/Disk performance and features
inferior memory management
Overall, I don’t want to do business or help in the success in an organization I do not like by offering up my data, watching their ads, and using their products less than necessary. I like some of the things Bill Gates has done, but it doesn’t change any of my views on this.
I felt like clarifying that the updates issues I faced were the last straw and that if anyone was interested, I listed the other reasons I quit working with them and never looked back. That’s why I wrote all that at the bottom.
Even if Microsoft does some things right, they still have a history of doing things wrong and have a bevy of other dark patterns. I do not trust them to get it right anymore. They could go back to their old ways tomorrow and I wouldn’t be surprised. Thankfully, it’s not my problem except at work
Win 11 Pro user here. It doesn’t care what time you set for updates, it’ll do them when it feels like anyway, or annoy the piss out of you with notifications.
It might be that I don’t leave the PC on all the time, I just hit sleep. But still, it shouldn’t strong arm me into updating after a day or two of the download. Also hate having to RegEdit Edge off the thing after each one.
Windows lets you pause updates for some time, maybe a week or so, after that you’re going to take them whether you like it or not. Granted, you had a week or so to prepare, so it’s ok to some extent, but don’t tell me Windows doesn’t force you…
Hmm, good to know, I’ll have to try, just out of curiosity. Is that available on Windows Home or just Pro? Anyways, it’s not something that many people would easily figure out, so for most non-technical people they effectively cannot disable them.
you still need to reboot your linux machine or relogin if you updated a process that’s currently running (and in most cases most system processes can’t be just restarted) (…and otherwise you’ll just stay at the old version bit with new data which might cause some instability)
yes, there’s kernel hot-patching but it only affects the kernel, only viable for minor and security upgrades, does not come pre-configured on most consumer distros and not really suitable for home use.
I remember some years ago there was a “malware” going around that would flash OpenWRT onto people’s routers, and set them to have more secure default settings.
There should be another thing like that, but one that upgrades Windows into a Linux distro.
When did you last use windows, lol? Windows is pretty damn stable nowadays. I don’t think an update has ever broken my windows 10 install that is still going from 2016.
I’ve gotten a number of calls from clients recently where a Windows update uninstalled the Bluetooth drivers, making their Bluetooth mouse and keyboard unusable.
I’ve even had a few where an update uninstalled the WiFi drivers so they couldn’t even download the drivers without a wired network connection.
Windows 10 & 11comes pre-packaged with generic wifi and bluetooth drivers that work with the vast majority of the common chipsets.
If a device has forgotten which driver it has, re-aasining the generic driver should be enough to get you operational enough to go grab any advanced drivers for extended device functionality.
Also, as an FYI, I had a fleet (~150) of decommissioned machines (probabaly 20-30 different model over 5 makes) I was converting into a Linux(Deb) distrubuted node automation farm. The amount of times I had to go find drivers (network interfaces were the cost common) that supported the hardware that Linux didn’t have default driver support for was prevelant. That was a very long 2 weeks.
I run arch BTW, 7 years throwing it down stairs, running commands that I had no idea what they did, learned linux from scratch deleting chunks of my hdd compiling and installing random software, never once had it break bad enough to reinstall . I bet you love ltt too haha… maby you should stick to a beginner os like Windows, I’ve heard Apple is even easier… or why don’t you just pay someone smarter than you to host and troubleshoot your os while they market your info and habits to the highest bidder… oh wait
Been running Arch on my work laptop for over a year. Still waiting for the fabled difficulty and update breaks. Starting to think in modern times its perpetuated to keep people on Windows.
Must be nice. It’s been about seven years since I last dove into Linux, so maybe things have changed. But also in that time, windows became even more stable than it was, and it’s silky smooth these days.
I don’t see any benefits to even trying Linux again.
The user. Depending on what they try to do, it can easily break Linux. (looking at me somehow breaking KDE Plasma and somehow fixing it without understanding how it broke or how I fixed it)
Updating (from what I understand, mostly a big issue on rolling release distros like Arch or Manjaro). Bleeding edge software with major bugs the stable release don’t get can always cause instability.
Though, I will say, that I’ve never had win10 crash on me unless I have too much stuff open or am being an absolute idiot. Windows always seems to be stable, at least I’ve never had issues for a long time.
Let’s be honest though. I’m a big fan of Linux/Unix systems, but if (not saying that’s necessarily the case) a normal user can break their installation by being a normal user, it’s not suited for normal users.
Windows is a pain in the ass imo, but pretty hard for a normal user to break in my experience.
Well, the US government has at least twice broken into infected US devices and fixed things. IDK about installing OpenWRT but the stories have some overlap
Sometimes people just don’t think about that people can have different wants and needs.
All, literally every game I want to play runs great in Linux, and my hobbies of self hosting, development, homelabbing, and data hoarding are all leagues better on it.
That doesn’t make a good choice for my friend that only logs on to play destiny 2. It also doesn’t matter why, to my friend, its a bad choice. It could be the devs are chained and lashed by Microsoft for even mentioning Linux in the office, but what matters to someonethatt only wants to play that game with friends is whether it works.
I’ve been using Linux desktop for a good 20 years now. All debian based distros (loads of them) do, all redhead based ones do, and those two together likely comprise the majority of distros.
I can’t remember the last time I rebooted my desktop (or servers, for what it matters) beyond a power outage in the office
Your updates both do not apply kernel updates but also aren’t applying in general unless you are restarting all apps, services, and sessions. Basically just reboot.
Only servers administrated well do online updates correctly.
Update small system components (packages) and load the old into ram until rebooting; I don’t think this is possible on windows.
A/B Image Based Updating; Android and a few Linux distros have this; probably one of the most stable methods.
Live boot updates/Kernel-space Hot Patching; found mostly in Linux servers, and distros with a patched kernel; used mostly for security updates which is what windows is doing here, but Linux can do feature updates this way too.
Windows is very lazy about reboots. Minesweeper changed? Better reboot.
Chrome also got infected with this laziness. It used to be that you had to restart chrome once a month, now it’s almost every day. Among many other reasons, that’s why I’m happy to be using Firefox again.
The chrome OS is method is pretty cool having a mirrored partitions the one not being used gets updated if there’s an error the other one gets booted and reverted
Had a movie stop playing the other week (I use my PC as a Jellyfin server and watch on a Nvidia Shield in another room). I thought something had crashed, but when I went upstairs to check, it had realised nobody was watching it and fucking rebooted.
Steam itself does support VR on Linux, but most of the actual hardware (like Meta headsets) don’t have drivers for Linux. The ones that do (Valve Index) are buggy, but not unusable. But even then it doesn’t get you far, because 90% of VR games won’t run on Linux, even with Proton.
So Steam is not the problem. Hardware support and developer support is the problem. Can’t really blame developers for not caring, even if they make their VR game work on Linux almost no one would be able to play it anyway, so why bother. It won’t get anywhere unless hardware manufactures start making actual drivers for their headsets on Linux. Meta practically controls the market and they don’t care, so here we are.
A Steamlink app was added to the Meta store recently. It supposedly allows playing streamed desktop VR. I have been meaning to try it with Steam on my Linux desktop, so I can’t really vouch for it yet, it could just not work. And who knows if Proton works for any specific VR games.
I actually meant Group Policy Editor. Sometimes I make mistakes like that. I will not dive into how precisely I made the mistake.
Coming to your second point, of course it is vulnerable, but I meant it in a practical sense. I am not here to waste time debating, so I am leaving it at that.
Honestly not being able to move the start bar and being told it won’t be changed because their awful new start menu needs it that way was a dealbreaker. Been running Linux Mint exclusively on my desktop for the past few months and it’s been pretty smooth, even for playing games. Thank goodness for Proton!
Yup. Been using Linux as my primary desktop for years, I think I switched back to windows 2012-2015 or something, then I came back ever since. More and more games are using tools that are cross platform now too - like unity for example. I only imagine compatibility getting better. The installation experience has been better since live CDs were a thing too which is hilarious since windows still has a terrible install UI.
I’ve been using both OSs for over 20 years and the ONLY reason I use windows is for CAD (just 2d). All the foss options have potential but are very poor options for a longtime autocad user. Wine implementation is currently broken/terrible. VM is sorta a fallback option but doesn’t run as fast as a native windows machine.
I plan on switching to Librecad or something similar but it’s like a 10/20 year plan and something tells me I’ll have to develop the features I want myself.
I don’t think those words describe what the intended behavior is, no. I think it’s supposed to be seamless and not really too noticeable. That’s the impression I got from the article anyway.
windowscentral.com
Active