> Lack of familiarity with AI PCs leads to what the study describes as "misconceptions," which include the following: 44 percent of respondents believe AI PCs are a gimmick or futuristic; 53 percent believe AI PCs are only for creative or technical professionals; 86 percent are concerned about the privacy and security of their data when using an AI PC; and 17 percent believe AI PCs are not secure or regulated.
Is being concerned about your privacy really a misconception here?
What would be the point of an "AI PC" if not to run models locally? I'm very much uncomfortable with sending my keystrokes (or my codebase) off to some remote server. If I can run the model locally, the privacy problems in theory vanish and I'm much more likely to use the tech.
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!
Tons of software runs locally but then still exfiltrates your data for a variety of reasons. Ads, product improvement, metrics, cross-device syncing, etc.
Access to the latest foundation models, which can’t be run locally. AI feels like it’s in this really weird place where the latest Claude model sets expectations that can’t be matched by an on-device model.
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”
> What would be the point of an "AI PC" if not to run models locally?
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>
There's not really a standard definition of an AI PC, but if there's anything beyond marketing, it would take the form of a neural processing unit, which supposedly makes some AI-related task fast by providing hardware acceleration. Cited examples include voice recognition and "training models", although I'm pretty sceptical about the later being at any significant scale.
I suspect what it will actually boil down to in many cases is having recall enabled by default and an always-listening voice assistant.
It's also not a misconception that "AI" PCs are for creative/technical professionals when the vast majority of the marketing is about shitting out creative(?) or technical(?) works(?) to reduce and ultimately democratize that workforce.
It shouldn't, one of my friends is a professional copywriter and she's very pro-AI because it raises the writing bar for the devs she works with. "I can write better than AI, but AI can write better than you, use AI" is her go-to line with her coworkers.
I was looking at laptops recently, and I noticed that the marketing blurbs on product pages are really getting extreme with the AI stuff in a weird SEO-like way. For example, this is on Costco's page for an Acer:
> 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.
I find the concept of "AI PC" to be somewhat nonsensical in the absence of a definition that is about the hardware.
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.
Assuming AI means LLM, at this stage I've come across two broad categories of implementation that are actually interesting and useful to me as a user.
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...
I think we just have a tendency to anthropomorphize things that are a bit too complicated to understand fully. Like a child calling the clutch mechanism in a yo-yo a "brain" for example. It's not that the yo-yo can really think, it's just that it has a behavior that seems that way. So indeed once you've upgraded your system to the point of not fully understanding how something that isn't a human could achieve whatever emergent behavior occurs, go ahead and anthropomorphize it by calling it AI.
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.
Televisions have been fairly egregious in recent decades. Curved and 3D the primary suspects. Unfortunately "smart" turned out to be a way to make money despite the protests of the consumers so that isn't likely to be phased out/fixed unless regulation separates them from their ill-gotten partnership revenue.
This is a natural phase of major tech breakthroughs. It is not a negative indicator. The opposite.
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.
> The chipmaker, which is quite keen to see people buy the AI PCs sold by its hardware partners
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.
Commercial software and ignorance make users less productive... I do not spend such time in finding my files thanks to Emacs/org-mode/org-attach to master them, I do not waste time with Office suite to prepare documents, I use LaTeX and org-mode and that's hyper faster once learned.
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.
> Commercial software and ignorance make users less productive...
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.
I suggest https://youtu.be/5yy6XvuO2aM summarized to "we want LLMs as life companion of our customers so we can drive their choices an LLM user interaction at a time". Of course such model can't work much with skilled customers.
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.
The reality is AI is largely in the uncanny valley right now. This is great because it weeds out the noise and gives the focused long term builders breathing room.
> Lack of familiarity with AI PCs leads to what the study describes as "misconceptions," which include the following: 44 percent of respondents believe AI PCs are a gimmick or futuristic; 53 percent believe AI PCs are only for creative or technical professionals; 86 percent are concerned about the privacy and security of their data when using an AI PC; and 17 percent believe AI PCs are not secure or regulated.
Is being concerned about your privacy really a misconception here?
What would be the point of an "AI PC" if not to run models locally? I'm very much uncomfortable with sending my keystrokes (or my codebase) off to some remote server. If I can run the model locally, the privacy problems in theory vanish and I'm much more likely to use the tech.
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!
Tons of software runs locally but then still exfiltrates your data for a variety of reasons. Ads, product improvement, metrics, cross-device syncing, etc.
Access to the latest foundation models, which can’t be run locally. AI feels like it’s in this really weird place where the latest Claude model sets expectations that can’t be matched by an on-device model.
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”
Looking at the logs for Apple Intelligence on my iPhone 15 Pro Max, almost everything is actually run remotely
What is not run locally?
> if the AI PC is just a regular PC with a cloud bot integrated, then ... what even is the point?
To trick people into buying new hardware, lest they get left behind in the AI race
> What would be the point of an "AI PC" if not to run models locally?
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
Do the AI PCs do the work locally or transmit everything remotely for processing like chatgpt?
There's not really a standard definition of an AI PC, but if there's anything beyond marketing, it would take the form of a neural processing unit, which supposedly makes some AI-related task fast by providing hardware acceleration. Cited examples include voice recognition and "training models", although I'm pretty sceptical about the later being at any significant scale.
I suspect what it will actually boil down to in many cases is having recall enabled by default and an always-listening voice assistant.
It's propaganda, as the article points out: "The chipmaker, which is quite keen to see people buy the AI PCs sold by its hardware partners,"
It's also not a misconception that "AI" PCs are for creative/technical professionals when the vast majority of the marketing is about shitting out creative(?) or technical(?) works(?) to reduce and ultimately democratize that workforce.
"Democratize" used to have positive connotations, as in "democratize access to information". I didn't anticipate that it would have negative ones.
It shouldn't, one of my friends is a professional copywriter and she's very pro-AI because it raises the writing bar for the devs she works with. "I can write better than AI, but AI can write better than you, use AI" is her go-to line with her coworkers.
When everyone is special noone is, as the old saying goes.
There's nothing herently good or bad about everyone getting access to or becoming capable of something.
I was looking at laptops recently, and I noticed that the marketing blurbs on product pages are really getting extreme with the AI stuff in a weird SEO-like way. For example, this is on Costco's page for an Acer:
> 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.
There is absolutely zero meaning that can be derived from that marketing blurb.
"This transportarion vehicle incorporates four toroid acceleration devices plus one toroid guidance machine."
I.e. the car has four tires and a steering wheel.
> pages are really getting extreme with the AI stuff in a weird SEO-like way
you mean in a weird generative AI way?
"Certainly. Here is a new blurb that promotes the product while using the term AI four additional times."
Jesus christ. Someone's getting fucking piledrived... six times!
https://ludic.mataroa.blog/blog/i-will-fucking-piledrive-you...
Ironic chatgpt's summary of the article
https://chatgpt.com/share/6743194c-ae00-8001-af3c-0915007579...
I find the concept of "AI PC" to be somewhat nonsensical in the absence of a definition that is about the hardware.
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.
In the 90's there were "Multimedia PCs"
https://en.wikipedia.org/wiki/Multimedia_PC
Assuming AI means LLM, at this stage I've come across two broad categories of implementation that are actually interesting and useful to me as a user.
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...
I think we just have a tendency to anthropomorphize things that are a bit too complicated to understand fully. Like a child calling the clutch mechanism in a yo-yo a "brain" for example. It's not that the yo-yo can really think, it's just that it has a behavior that seems that way. So indeed once you've upgraded your system to the point of not fully understanding how something that isn't a human could achieve whatever emergent behavior occurs, go ahead and anthropomorphize it by calling it AI.
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.
> Our role as technology leaders is to support this transition to AI-assisted living
Why is it that we all woke up one morning and every corporation is suddenly saying this same thing?
Their execs are probably all subscribed to the same corporate leadership magazines?
Just as with 3D televisions, be patient and this trend will pass.
Televisions have been fairly egregious in recent decades. Curved and 3D the primary suspects. Unfortunately "smart" turned out to be a way to make money despite the protests of the consumers so that isn't likely to be phased out/fixed unless regulation separates them from their ill-gotten partnership revenue.
This is a natural phase of major tech breakthroughs. It is not a negative indicator. The opposite.
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.
> The chipmaker, which is quite keen to see people buy the AI PCs sold by its hardware partners
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.
I am not an AI booster, but I would expect a learning curve slowing down current tasks for pretty much any technology that needs to be learned.
> AI PCs make users less productive
Idiot discovers that more "tools" make users less productive.
Commercial software and ignorance make users less productive... I do not spend such time in finding my files thanks to Emacs/org-mode/org-attach to master them, I do not waste time with Office suite to prepare documents, I use LaTeX and org-mode and that's hyper faster once learned.
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.
> Commercial software and ignorance make users less productive...
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.
I suggest https://youtu.be/5yy6XvuO2aM summarized to "we want LLMs as life companion of our customers so we can drive their choices an LLM user interaction at a time". Of course such model can't work much with skilled customers.
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.
The reality is AI is largely in the uncanny valley right now. This is great because it weeds out the noise and gives the focused long term builders breathing room.