AI PCs make users less productive
theregister.com
theregister.com
Is being concerned about your privacy really a misconception here?
If folks don't understand that, then yes I'd say it's a pretty big misconception and needs to be better marketed as a key feature.
And if the AI PC is just a regular PC with a cloud bot integrated, then ... what even is the point? You can already do the remote chatbot thing with a regular PC, privacy nightmares included!
Even Apple Intelligence is getting a lot of negative feedback by the review crowd due to its limitations like being unable to summarize a very large document (which is pretty much the point of such a feature).
The problem is that AI has few well-defined use cases and a mountain of expectations, and this really shows in the execution by these companies. It’s hard to build good products when the requirements are “we don’t really know”
Unrealistic for now because running ML is slow but even if yes, even laypeople know by now you can break an LLM to do what it isn't supposed to. Since this LLM has unlimited access to your personal data to be useful, if I get to it I don't even need to bypass any secure enclaves or what not because it will tell me things I ask for in plain <insert your language>
All eggs in one basket
To trick people into buying new hardware, lest they get left behind in the AI race
I suspect what it will actually boil down to in many cases is having recall enabled by default and an always-listening voice assistant.
Instead, centralized AI tools will be creating more of our music for us, which is the exact opposite of democratization.
What level of mental gymnastics is this? This is like saying that cars enable people to not travel.
Music is in the ears of the beholder, it doesn't matter how it was produced. If a tool lets more people compose music, which "AI" does, then it's democratizing composing music.
Likewise: Crossbows and then guns democratized violence. Cars democratized personal transport. Currency democratized procurement. The printing press democratized writing. The internet democratized information.
Democratizing is about making something more accessible to the commons. Whether that will be good or bad is not the concern of democracy.
No, sitting in a car that's traveling, you are actually traveling with it regardless of whether you are driving it, so it's entirely unlike saying that cars enable people not to travel.
It's more like saying that giving everyone access to a private robot driver doesn't somehow democratize driving and instead enables people not to drive. You can instruct the driver to drive where you want to go on your behalf, and you can travel along with them in the car, but it hasn't somehow enabled anyone but the robot to drive. It has however enabled you not to drive: even in situations where you would otherwise have driven you can now rely on your robot driver to do it for you.
What the robot driver has democratized is access to something that can drive for you, not driving. Similarly, what AI/ML/LLM/whatever-based music generators and the like have democratized is access to something that can create music for you, not creating music.
Not even the ATC can be considered to be pilots simply for giving instructions to others who fly planes, even though they give pilots much more detailed and involved instructions than anyone gives e.g. Suno.
> Music is in the ears of the beholder, it doesn't matter how it was produced.
Even if we assume that this is true, it has no bearing on my argument: whether or not it matters how music was produced is irrelevant to the question of who or what created the music.
> If a tool lets more people compose music, which "AI" does, then it's democratizing composing music.
But it doesn't let more people compose music. The "AI", for example Suno, creates the music for you, and your input is more akin to the instructions you give to the driver: extremely high level and entirely removed from the work involved in producing the result. Similarly, as a composer, writing a piece that performing artists then interpret and use as instructions for a performance doesn't somehow make me a performing artist, despite your instructions being much more detailed than any instruction that has ever been given to Suno, and therefore enabling the composer to be more involved in the outcome of the performance.
> Likewise: Crossbows and then guns democratized violence.
Will you agree that the degree of involvement has any bearing on the judgement whether it is something I do or it's something others do for me? For example, that having private robot driver driving me around at my behest doesn't make me a driver myself, but aiming and firing a gun at someone with the intent to cause violence to them makes me violent, despite the fact that I'm not the bullet that ultimately caused the harm?
If so, the reductionist argument is misleading and we have to consider the degree to which you are involved in the process of creation when you tell Suno to produce a song.
> Whether that will be good or bad is not the concern of democracy.
Whether it's good or bad is also completely irrelevant to my argument, and seemingly also to your response to it, so why bring it up at all?
An "AI" is a tool just like a car or a crossbow, they enable its user to do something more easily. Cars democratize travel, crossbows democratize violence, "AI"s democratize productivity.
>Whether it's good or bad is also completely irrelevant to my argument,
You literally replied to a comment concerning the possibility of the word "democratize" having innate positive and/or negative meanings.
I'm certainly don't attribute any intent to anything in my reasoning, but yes, most tools have a purpose. Suno's purpose, for example, is clearly to create music based on natural language prompts. Not that I have expressed that before, or that my argument relies on that fact, so whether or not you agree, it's irrelevant to what I'm saying.
> Additionally, you might also be attributing a persona to a static and mindless tool which is also nonsense.
I might? You should respond to my argument, not address things you baselessly think I believe, which is a waste of both our time.
> You literally replied to a comment concerning the possibility of the word "democratize" having innate positive and/or negative meanings.
I replied to say that "democratize" is the wrong word. My argument to that end doesn't hinge on whether democratization is a positive thing or a negative one. This discussion would be much easier if you responded to my points directly instead of trying to mischaracterize my argument as something it is not. Your response here is very light on directly addressing anything I said, and heavy on insisting that I believe things that I don't.
I can't because you're talking nonsense.
You are essentially saying that guns commit murder, not the user who shot the gun. You are saying that the tool carries intent and purpose, not the user. It's ridiculous, faulty logic and I cannot respond to that other than to point out it's quackery.
You can't, because your argument at this point doesn't pertain at all to my argument.
> You are essentially saying that guns commit murder, not the user who shot the gun.
I am explicitly, literally saying that it's not, referring to that exact example, which you'd know if you'd actually taken what I've written into account when writing your response.
> You are saying that the tool carries intent and purpose, not the user.
Again, I am not, but please go ahead and point out where you believe I am.
> It's ridiculous, faulty logic and I cannot respond to that other than to point out it's quackery.
That sounds like a you problem. I have no problems responding directly to your points despite of your failure to address my argument at all, to point that fact out if nothing else. This is one of the perks of intellectual honesty: I don't have to pretend that I can't make sense of what you are saying to argue for my point.
Are you saying that programmers don't create software because they just "prompt a compiler"? Are you saying authors don't write books because they type into a word processor? Are you saying illustrators and artists don't draw because they are using a tablet? Are you saying construction workers aren't constructing because they're using excavators and cranes?
I appreciate that many people suffer from AI Derangement Syndrome, but it's a ridiculous notion. "AI" is a tool like any other tool.
Asking some gen AI to create music is as much of a creative effort as ordering dinner at a restaurant. Are you a chef when you do that?
There's nothing herently good or bad about everyone getting access to or becoming capable of something.
> AI Ready for Tomorrow
> Ready for the ever-evolving possibilities of AI? This Swift Go AI PC integrates Intel’s new dedicated AI engine—Intel® AI Boost—with Acer’s own AI solutions, for more intuitive and enjoyable AI experiences.
It really seems like some kind of stupid joke.
I.e. the car has four tires and a steering wheel.
you mean in a weird generative AI way?
https://ludic.mataroa.blog/blog/i-will-fucking-piledrive-you...
https://chatgpt.com/share/6743194c-ae00-8001-af3c-0915007579...
Just working out the age of my personal desktop computer has a Ship of Theseus problem - but safe to say 20+ years. However, it now has a graphics card with an RTX 3060 with 12 GB of GPU, and NVMe SSDs, and can run inference on 7B parameter 4 bit quantised Transformer LLMs, and generate images with large diffusion models. I've also used it for many applications that would count as AI before the latest generative AI hype cycle.
So is it an AI PC? At what point did it become an AI PC? Or is a self-built machine in which you swap parts inherently never an AI PC?
Given the fact that it is so amorphously defined, I would consider the term to be purely marketing fluff.
1) A box on the screen where I can chat with one to do ideation or really anything I want.
2) A command-driven approach where I hit a hotkey, type a prompt and the response is dumped out in front of me, possibly I had some text selected which heavily influences the response.
These are both pretty cool tbh and developers will have a field day for years finding sensible ways to incorporate them into programs.
None of this has anything to do with driving the hardware upgrade cycle since most of the models are running in the cloud. But driving hardware upgrades is what these marketing people are really trying to do when they talk about AI PC. They are irrelevant people, but they need to convert everything they see into a reason to buy a new PC.
That's what they get paid for. Monkey marketer see trend, monkey marketer do marketing. Monkey steal your attention.
Maybe a LLM will replace THEM soon. After all it's basically a digital version of the million monkeys on typewriters...
There's not a specific line in the sand, although tasking it with machine learning (in which outcomes improve based on collecting runtime inputs, rather than based only on its creator adding capabilities) would be a decent one. That's fairly human-like, while non-ML workloads are more plant-like.
It is normal for most tech to have an exploratory period, where its potential is clear, but its immediate economic impact is negative during iterations of product-market fit adaptation.
Normally, the producer of new tech eats most or all of the risk and cost of the search for product-market fit.
But some tech is so compelling, that customers feel the strategic need to participate in the discovery loop too.
Obviously, there are upfront costs and risk deploying/trying tech that is still hit-and-miss. But during a sea change, there is also risk in not experimenting and adopting/adapting early.
Why is it that we all woke up one morning and every corporation is suddenly saying this same thing?
Hah, no they very much aren't, as Intel are an insignificant player at the moment. As long as Intel is as far behind as they are, they'd rather overall investment in hardware goes down. The second Intel comes out with a leading chip, you'll suddenly see them come up with a study with the opposite result.
Essentially FLOSS tools must be learned in time, typically at school, than you can profit for life, commercial software is an endless low-learning process full of frustration to do anything. No LLM can solve that, it's a chosen design for business purpose.
This is exactly what I have observed too. Notice how all commercial software comes very polished and modern with a very superficial UI. Despite most software products being hundreds of times more complex than simple electronic gadgets, such gadgets often come with more complex manuals.
It's obvious that the true end-game of AI is not to make end users more productive, but rather de-skill end-users and make them entirely beholden to third-party software. Most companies would love the tradeoff of making their employees more like cogs at the cost of lower productivity.
Alas, I too am held hostage by Emacs as most software doesn't even come close to the flexibility it provides.
In the end is the same plot in every time, in 1894 an Italian Education Minister for SEVEN times, Guido Baccelli, stated "we need to teach just to read and write, be careful teaching history, we MUST set aside anti-dogmatism and critical thinking, the people MUST NOT THINK or we will be in trouble".
We see the same trend everywhere, in every profession as well, even in education https://www.theatlantic.com/ideas/archive/2020/08/i-was-usef... the issue is that such people are unable to innovate and smart and educated people are always scarce so if you do not cultivate knowledge you simply lost it, knowledge is the sole "natural resources" that grow with use instead of the contrary.
Idiot discovers that more "tools" make users less productive.