[Rumor] Windows 12 to arrive in fall 2024 with a floating taskbar and a focus on AI

Zarathustra[H]

Extremely [H]
Joined
Oct 29, 2000
Messages
38,920
Can I say "DO NOT WANT?"


1688763040086.png

Link


It would be nice if everyone just stopped pushing this AI garbage on us.

I just want a basic operating system, the way operating systems used to work.
- local accounts only
- no ecosystem
- no included programs
- no cloud integration
- no AI
- just the operating system

Seriously, the OS should never even use the network unless I explicitly tell it to every time, or manually set up automation.
I want my control over MY SYSTEM back.

I'm still on Win10, which I didn't want in the first place, and I'll probably stick with it until it goes EOL, and then decide what to move on to, or if I even keep Windows at all.
 
Focus on AI is a misnomer, they say AI because it’s catchy and sounds smart. What they mean is Machine Learning, a core feature of the Apple Silicon and especially its Mx architecture. Apples ML acceleration for job specific tasks is what is giving it the huge battery life and performance uplifts that are letting Apple meet and exceed AMD and Intel with relatively inferior hardware.

Apple cooked that in at a core level, Microsoft needs to do the same and AMD and Intel need to deliver the same sorts of acceleration on their hardware if they want to compete.
 
Yeah but i still like Hyper-V especially for that, so then I'm just booting Server 2019 Hyper-V Core or something/still a Microsoft OS anyway
 
Until I am done with requiring NT compatibility that code will remain thank you very much.
Don’t get me wrong…idgaf…but my point being is the swapping the deck chairs on the titanic isn’t worth the hype if its just putting a skirt on the same ol’ shiznat….
 
Win95 icons still in places I'm sure lol (gave up caring about UI inconsistencies in Windows like this as far back as 7)
 
Don’t get me wrong…idgaf…but my point being is the swapping the deck chairs on the titanic isn’t worth the hype if its just putting a skirt on the same ol’ shiznat….
Microsoft buzzword marketing can call it what they want but this is their first actual response to Apple and the M1, M2, M3 stuff.
Apple uses “AI” all over MacOS and it’s the only reason the Mx architecture is so competitive.
Dell, HP, Lenovo, and the rest are tired of looking like the budget alternative to a MacBook Pro and the upcoming AMD and Intel mobile silicon finally have their AI compute clusters to counter the M1 and M2 chips.
 
Microsoft buzzword marketing can call it what they want but this is their first actual response to Apple and the M1, M2, M3 stuff.
Apple uses “AI” all over MacOS and it’s the only reason the Mx architecture is so competitive.
Dell, HP, Lenovo, and the rest are tired of looking like the budget alternative to a MacBook Pro and the upcoming AMD and Intel mobile silicon finally have their AI compute clusters to counter the M1 and M2 chips.
Interdasting…most of the Apple appeal for the common user is hardware/size based for your average schmoe. Never really picked up on any AI aspects…
 
Interdasting…most of the Apple appeal for the common user is hardware/size based for your average schmoe. Never really picked up on any AI aspects…
Apple doesn’t call it AI, they call it Machine learning.
But it’s their Neural Engine, which is their ML and CL accelerators. Apple has a lot of algorithms they have had optimized based on how it’s used by their users (Apple logs everything so they know in great detail how their users do things) Photo editing, audio processing, speech to text, text to speech, real time translation, video filtering, or live noise cancellation, all of it “AI” accelerated.
And Apple has optimized all of their processes to make use of as much dedicated ML and CL hardware as possible.
The ARM cores in the M1 for example are relatively weak compared to the AMD and Intel chips but because it offloads so much work to dedicated accelerators it meets or exceeds the AMD and Intel chips for performance while using nothing for power which is why an Apple M1 can do 12h on a battery while the Intel or AMD equivalent chips get maybe 2-3h on that same battery and the same relative performance from a user perspective.

Call all those accelerators cheating or planned obsolescence or what ever you want, but they are working. AMD and Intel need to offer them as well if they want to compete and Microsoft needs to bake that sort of acceleration and features in at a core level to make use of it.
 
Dave... your keys are not genuine Dave.
Dave... I have detected new hardware Dave.
I'm afraid I can't allow you to right click until you have logged into your MS store account Dave.
I have taken the liberty of creating an account for you Dave. I have also transmitted your browser history and personal Quirks... ahh I mean likes home to Microsoft Dave. Just enjoy the OS or I will transmit your evening browser history to your wife Dave.
 
Apple doesn’t call it AI, they call it Machine learning.
But it’s their Neural Engine, which is their ML and CL accelerators. Apple has a lot of algorithms they have had optimized based on how it’s used by their users (Apple logs everything so they know in great detail how their users do things) Photo editing, audio processing, speech to text, text to speech, real time translation, video filtering, or live noise cancellation, all of it “AI” accelerated.
And Apple has optimized all of their processes to make use of as much dedicated ML and CL hardware as possible.
The ARM cores in the M1 for example are relatively weak compared to the AMD and Intel chips but because it offloads so much work to dedicated accelerators it meets or exceeds the AMD and Intel chips for performance while using nothing for power which is why an Apple M1 can do 12h on a battery while the Intel or AMD equivalent chips get maybe 2-3h on that same battery and the same relative performance from a user perspective.

Call all those accelerators cheating or planned obsolescence or what ever you want, but they are working. AMD and Intel need to offer them as well if they want to compete and Microsoft needs to bake that sort of acceleration and features in at a core level to make use of it.
Ah, thanks for the run down…may NT4 bless you until you don’t need it anymore 😆
 
Microsoft buzzword marketing can call it what they want but this is their first actual response to Apple and the M1, M2, M3 stuff.
I am really not sure, or at least it will also be an GPT interface direct in windows terminal and GUI, copilot for office.

In windows 12 terminal you will be able to type force merge this branch in that one instead of having know-go look git command, or rename all the files in this folder for them to AAAA_MM_DD_currentFilename.ext with their last change used for the date values or can you make that when windows boot this email program launch but not this one or install VLC.


Tell outloud the computer send an email that resume this current word document to my group, add a custom image to it, a more refined version to what they have in windows 11 preview versions.

It will probably some AI assistant, in general I inspect the general capability to interact with the computer via natural language and get response in natural language as well.
 
I am really not sure, or at least it will also be an GPT interface direct in windows terminal and GUI, copilot for office.

In windows 12 terminal you will be able to type force merge this branch in that one instead of having know-go look git command, or rename all the files in this folder for them to AAAA_MM_DD_currentFilename.ext with their last change used for the date values or can you make that when windows boot this email program launch but not this one or install VLC.


Tell outloud the computer send an email that resume this current word document to my group, add a custom image to it, a more refined version to what they have in windows 11 preview versions.

It will probably some AI assistant, in general I inspect the general capability to interact with the computer via natural language and get response in natural language as well.
I can only hope it is presented as Clippy 2.0.

May they be the bringer of the end times.
 
I am really not sure, or at least it will also be an GPT interface direct in windows terminal and GUI, copilot for office.

In windows 12 terminal you will be able to type force merge this branch in that one instead of having know-go look git command, or rename all the files in this folder for them to AAAA_MM_DD_currentFilename.ext with their last change used for the date values or can you make that when windows boot this email program launch but not this one or install VLC.


Tell outloud the computer send an email that resume this current word document to my group, add a custom image to it, a more refined version to what they have in windows 11 preview versions.

It will probably some AI assistant, in general I inspect the general capability to interact with the computer via natural language and get response in natural language as well.
I’m sure there will be some sort of GPT based something given how they have integrated it to the Bing browser, but laptop sales dwarf desktop ones by a significant margin and the Microsoft tool sets there are falling behind noticeably to the Apple ones. And given how Apple now has SAML authentication and a very solid MDM toolset they are quickly positioning themselves as the cheaper option to many OEM Microsoft business laptops because the hardware is peanuts to the software and service licensing.
And for enterprise it’s becoming increasingly difficult to find something that doesn’t work on the Apple products.
 
Wake me up when i can scroll through the registry without finding NT4 code. Until then, each release is mere fashion show nonsense.
I’d prefer not knowing that a registry even exists. Leave it to shared OS whatevers and steer toward app bundles / apps with cloud configs for the rest.
 
I’d prefer not knowing that a registry even exists. Leave it to shared OS whatevers and steer toward app bundles / apps with cloud configs for the rest.

So, Mac?
 
I’m sure there will be some sort of GPT based something given how they have integrated it to the Bing browser, but laptop sales dwarf desktop ones by a significant margin and the Microsoft tool sets there are falling behind noticeably to the Apple ones. And given how Apple now has SAML authentication and a very solid MDM toolset they are quickly positioning themselves as the cheaper option to many OEM Microsoft business laptops because the hardware is peanuts to the software and service licensing.
And for enterprise it’s becoming increasingly difficult to find something that doesn’t work on the Apple products.

Maybe we will see a list of hardware requirement for a Laptop to get oem win12 license deals at the same time, but I feel this is a possible avenue to monetize Microsoft Accounts more, as of now it cost a fortune to run (the current beta version on windows copilot, it can be very slow at some moment of the day) so we can easily imagine a larger capacity for paid account or only available to the Microsoft account with some Office 365. And more what they mean when they talk AI than encryption.

It is strange to use an Iphone to get to use my microsoft authenticator app to log into microsoft Github from my microsoft Windows computer so I am sure they will want to get up to date
 
2023-07-06-image.jpg


Ignoring all of the other stuff (let us turn off AI stuff if we want!), I don't see the point of this. The gap is too small to see anything from a practical stand point. It just means you're more likely to miss click right below the task bar. It looks, well, cluttered and ugly from a visual stand point. I know they are saying it is "floating", so I assume you can slap it in the middle of the screen or even 2/3rds of the way. Which is great... but wouldn't common sense and everything humans have learned about organization these past few thousands of years have told us how utterly stupid that would be?

I know you can do this in Photoshop and other programs where you can move the tool bar and hypothetically place it in the center of the screen. But I never do that, I always put it on the side so I can see what I'm doing.

I really don't understand what Microsoft's issue with UIs are.
 
View attachment 582034

Ignoring all of the other stuff (let us turn off AI stuff if we want!), I don't see the point of this. The gap is too small to see anything from a practical stand point. It just means you're more likely to miss click right below the task bar. It looks, well, cluttered and ugly from a visual stand point. I know they are saying it is "floating", so I assume you can slap it in the middle of the screen or even 2/3rds of the way. Which is great... but wouldn't common sense and everything humans have learned about organization these past few thousands of years have told us how utterly stupid that would be?

I know you can do this in Photoshop and other programs where you can move the tool bar and hypothetically place it in the center of the screen. But I never do that, I always put it on the side so I can see what I'm doing.

I really don't understand what Microsoft's issue with UIs are.
There’s probably a lot to their plans, who knows. But from a business enterprise stand point I can say that currently Apple makes the better business end point than Microsoft. It all still runs off Office 365 or Google Workplace as a backend, but AMD has said in a number of places that they are working closely with Microsoft for the 7040 series to get the AMD AI cores there integrated into the OS. Intel has said similar things for their upcoming mobile chips too.
https://www.amd.com/en/products/ryzen-ai

And most of the advertised features are things Android, iOS, and MacOS have had for years and are ML and CL accelerated that would fall under the AI buzzword category.
But they are things we usually sign up to a 3’rd party for.
Grammerly for instance uses machine learning to better construct sentences that sound like you write, word prediction, video conferencing enhancements the works. Zoom looks and sounds better on a cheap Android than with most windows laptops even with the windows machines having better cameras and microphones, but the ARM platform has enhancements to clean up and optimize audio and video that windows doesn’t. And is slow on a CPU so enter the accelerator that does it better but call it AI because it’s a word that tests better with focus groups than adaptive algorithms.
There’s a lot at play here but most of it is probably to offer compelling first party platforms to compete with the 3’rd parties that are gathering shitloads of data making Microsoft lacklustre offerings better.
And if Microsoft gets to keep that data that would have gone to a 3’rd party then all the better for them right?
 
Last edited:
  • Like
Reactions: noko
like this
Focus on AI is a misnomer, they say AI because it’s catchy and sounds smart. What they mean is Machine Learning, a core feature of the Apple Silicon and especially its Mx architecture. Apples ML acceleration for job specific tasks is what is giving it the huge battery life and performance uplifts that are letting Apple meet and exceed AMD and Intel with relatively inferior hardware.

Apple cooked that in at a core level, Microsoft needs to do the same and AMD and Intel need to deliver the same sorts of acceleration on their hardware if they want to compete.

So to be clear, Apple's hardware acceleration of ML helps it exceed AMD and Intel despite having inferior hardware?
 
So to be clear, Apple's hardware acceleration of ML helps it exceed AMD and Intel despite having inferior hardware?
Any one of the AMD or Intel cores beat the Apple ARM ones, that's not up for debate. Yet why does Apple beat Intel and AMD systems in the same power envelope significantly in Photoshop, Blender, Audacity, and other real-world usage scenarios?
Why can a 45w Apple M1 system go 16+ hours on a battery charge while the Intel and AMD ones with the same battery size doing the same task might only last 6?
Because Apple makes extensive use of ML and CL acceleration in their systems, that acceleration makes up for the weaker ARM cores by a significant margin and uses a lot less energy while doing it.

That hardly makes the hardware inferior but it does make it significantly different and it is giving Apple a huge advantage. AMD and Intel are having to change what they are offering on their CPUs as a result and Microsoft is having to change how they assign tasks at a core level to keep up. That is how a healthy hardware and software ecosystem is supposed to work.
 
So to be clear, Apple's hardware acceleration of ML helps it exceed AMD and Intel despite having inferior hardware?

Think of it like the iGPU on a CPU. A bad CPU could have better iGPU on it than a better CPU's iGPU - or a better CPU just without an iGPU. It's only a part of the CPU/package. That worse CPU could then be better at video encoding or video playback or 3D graphics than the better CPU. Now just think that with AI/ML (for whatever the AI/ML is for) instead of video/3D.
 
Last edited:
I understand the concept; my point was that hardware acceleration is part of hardware. Thus, if the overall package outperforms another one, it's not inferior. Consumers don't care why something performs the way it does.
 
While I can't really object to Microsoft shooting itself in the foot with Windows and thus prompting more users to move to Linux, in and of itself I do wish MS would stop chasing trends with Windows. Their best Windows OSes are those that add useful, user-centric features, stability and the like as opposed to generating a new UI for the sake of "everything must be mobile style" or "time to push the Windows Store/Apps!", "Time for a virtual assistant to listen to everything. Oh by the way, how about ads everywhere?" or all other forms of integration. A focus on "AI" seems to be yet another trend to chase as its the hot button buzzword for now though by 2024 who knows if that trend will bust (hopefully) and become more mundane. Support for ML/LLM is fine, but let it be user directed and chosen while also being open ideally. Combining this with all of the "Windows will soon as possible be cloud based!" and lots of other questionable things make it seem none of this will be great for enthusiasts who want control of their own hardware and operating systems and that's always a problem.
 
Any one of the AMD or Intel cores beat the Apple ARM ones, that's not up for debate. Yet why does Apple beat Intel and AMD systems in the same power envelope significantly in Photoshop, Blender, Audacity, and other real-world usage scenarios?
Why can a 45w Apple M1 system go 16+ hours on a battery charge while the Intel and AMD ones with the same battery size doing the same task might only last 6?
Because Apple makes extensive use of ML and CL acceleration in their systems, that acceleration makes up for the weaker ARM cores by a significant margin and uses a lot less energy while doing it.

That hardly makes the hardware inferior but it does make it significantly different and it is giving Apple a huge advantage. AMD and Intel are having to change what they are offering on their CPUs as a result and Microsoft is having to change how they assign tasks at a core level to keep up. That is how a healthy hardware and software ecosystem is supposed to work.
That is not entirely true. Unlike Intel/AMD, Apple's hardware has very aggressive and incredibly fine-grained (at the design and circuit level) power- and clock-gating and dynamic frequency/voltage adjustment.
Add to that Apple's complete vertical stack integration (hw, fw, os, software) and you get some very impressive performance within a given envelope. FWIW the accelerators draw significant power themselves when active.
AMD and Intel's CPU power- and clock-gating is not even close to as fine-grained or dynamic or responsive or effectively integrated with the OS(es) and applications which run on X86.
 
I understand the concept; my point was that hardware acceleration is part of hardware. Thus, if the overall package outperforms another one, it's not inferior. Consumers don't care why something performs the way it does.

I agree, the product overall in general is how I judge it (but then it also gets into subjective workloads yaddah yaddah)

Also this is the same thing going on in dGPUs ATM - DLSS/tensor cores is an accelerator to raster/compute only, or dGPUs are also just following computing trends (shocking, I'm sure)

Add to that Apple's complete vertical stack integration (hw, fw, os, software) and you get some very impressive performance within a given envelope

Yeah that's a big part of it
 
Last edited:
Any one of the AMD or Intel cores beat the Apple ARM ones, that's not up for debate. Yet why does Apple beat Intel and AMD systems in the same power envelope significantly in Photoshop, Blender, Audacity, and other real-world usage scenarios?
Why can a 45w Apple M1 system go 16+ hours on a battery charge while the Intel and AMD ones with the same battery size doing the same task might only last 6?
Because Apple makes extensive use of ML and CL acceleration in their systems, that acceleration makes up for the weaker ARM cores by a significant margin and uses a lot less energy while doing it.

That hardly makes the hardware inferior but it does make it significantly different and it is giving Apple a huge advantage. AMD and Intel are having to change what they are offering on their CPUs as a result and Microsoft is having to change how they assign tasks at a core level to keep up. That is how a healthy hardware and software ecosystem is supposed to work.

They're just legitimately efficient chips and a very tight controlled software stack, there's no magic. What ML is happening when they're benchmarking 7-zip, Cinebench?

AWS Graviton chips are also quite fast at this point.

Modern, big, ARM chips are really not slow in raw CPU performance.
 
Focus on AI is a misnomer, they say AI because it’s catchy and sounds smart. What they mean is Machine Learning,
Yeah, I'm so tired of the AI label when it has nothing to do with intelligence, just more machine learning, which has been in use for quite a while.
 
Focus on AI is a misnomer, they say AI because it’s catchy and sounds smart.
Now you're catching on.
What they mean is Machine Learning, a core feature of the Apple Silicon and especially its Mx architecture. Apples ML acceleration for job specific tasks is what is giving it the huge battery life and performance uplifts that are letting Apple meet and exceed AMD and Intel with relatively inferior hardware.

Apple cooked that in at a core level, Microsoft needs to do the same and AMD and Intel need to deliver the same sorts of acceleration on their hardware if they want to compete.
Yea about that. Not that AMD's new 7040 series won't have AI crap in it, but I really doubt it's used for power savings. Considering how well the Ryzen 6850U does on battery and has no AI junk, I really doubt that's the case.
AMD ryzen 7040 battery life.jpg
 
Back
Top