One-third of consumers reject AI on their devices, with most saying they simply don’t need it — latest report highlights privacy fears and potential costs among other real-world concerns

AI report shows scepticism among one-third of US consumers

(Image credit: Getty Images)

There is a disconnect between the enthusiasm for AI among the companies pushing its development and the general public at large, but it isn’t just scepticism that’s fueling it. In news that might not be entirely surprising, a new report by the consumer research firm Circana has found that, while most consumers now know what AI is, two-thirds of those opposed to it don’t believe it’s needed on their devices at all.

That’s despite a big push to put AI-powered capabilities onto our devices. For instance, the integration of Microsoft Copilot into Windows has continued apace, with new Windows laptops now shipping with a specialist Copilot key for quick access to its AI assistant. AI functionality has been spotted in almost every major consumer electronic device category available to date.

It isn’t just a concern about the functionality itself that is driving the cynicism. 43% of those who were negative about AI listed a worry about the costs of AI, not wanting to pay more for the extra functionality. That’s a concern shared by major banking firm J.P. Morgan last year, providing research that the costs of AI vastly outweigh returns, equivalent to a $35 payment for every iPhone user “in perpetuity.”

That's good news for the shareholders, but this survey shows that there is still work for AI companies to justify its widescale adoption. The range of issues – from privacy, to cost, to the actual usefulness of AI functionality, not to mention the environmental and ethical concerns – is likely to grow in intensity as the technology's impact on our lives continues to grow.

Follow 3DTested on Google News, or add us as a preferred source, to get our latest news, analysis, & reviews in your feeds.

27 Comments Comment from the forums

excalibur1814

CoPilot, this morning:
"Hey, how's it going?"
"It's okay, how,many 'Rs' are there in Strawberry?"
"Blah, blah, two!"
"That's wrong...", tries to end the chat, which was odd, "check the thesaurus or online, there's three!"
"Oh, yes, sorry I've made a mistake, how human of me, there's three in strawberry!"

By now, if Co-Pilot has been asked that question before (which it must have), one might assume that it would answer correctly.

What 'I' want, is Cortana to return and the ability to simply say, "Hey, Cortana, shutdown my PC!"
"Are you sure that you want to shutdown your PC?"
"Yes!"

Give me functionality.
Reply
EzzyB

How about those of us who simply don't trust it? We've seen, with Grok foremost, that AI will say whatever it's designer wants it to. Have you heard about the White Genocide in South Africa? (This in response to a question about a baseball pitcher's salary at a time that there were political questions about South African immigrants.)

People do not realize what AI is all about. Why are these companies spending trillions in an effort to develop it? All of them KNOW that there is going to be a huge financial fallout eventually and most of them are going to lose huge amounts.

Think about the most common. Do a Google search. If you haven't modified it in your browser like I have the first thing you see is what? The AI summary. Now how many people see that, read it and are satisfied with that answer? 8 out of 10? Over a couple of billion searches worldwide a day?

AI will tell you what it's designers want it to say. They function by the same basic law all computers have as long as they've existed. GIGO. Garbage In, Garbage Out.

The winner, after the bubble bursts, then controls information flow to billions. He who controls the AI, controls the narrative. That is, in fact, worth trillions to them.
Reply
Penzi

I remain very excited for the future of “AI” (I still put it in quotes occasionally, apologies) but am middling on interest in the present and snort in derision for the wasteful nature of the technology and the amount of money being burned in bonfires to its effigy.

As an adaptable rules-follower there is much to recommend it. As an image editor, rough code provider, and initial research assistant it’s already shown value. We’ll see where corporations attempt to take it and how users respond to those choice.
Reply
PEnns

It should be an option, not rammed down everybody's throat to please the billionaires' club of Tech Bros!

How is that not any different from Micro$haft's browser that nobody wanted, or even any M$ software foisted upon ALL consumers??
Reply
cknobman

The harder they push AI the more skeptical I am of it.

My life experience has taught me when something is being pushed on you, no matter how much you dont want it, there is always something nefarious behind it.
Reply
lazymangaka

It's an interesting technology that's deeply boring from an "average consumer" perspective. The vast majority of the things that these algorithms can do just isn't that useful for most people, at least in a direct "using the thing" kind of way. I get that we always need a "next big thing" to sell the new product, but this is a technology that would have been way better at just being implemented behind the scenes to make existing services better (voice assistants, image recognition, automatic captioning, image editing, to name a few) rather than being splashed around like the be-all and end-all of tech development.
Reply
timsSOFTWARE

The principles underlying LLM learning are really fascinating, and I think will be very useful in all sorts of ways in the future. But the rush to invest trillions in current-gen hardware just seems misguided for a bunch of different reasons. One, it's pushing toward a near-term economic crash. Two, the hardware won't even be cost-effective to run in 5-7 years' time.

Three, regardless of whether AI advances or stays the same, the huge datacenter, AI-in-the-cloud model probably won't work for commercial purposes long-term. Either the tech gets stuck for a while, and people want to run their LLMs on-prem/locally for privacy and security reasons, or continuous learning AI is developed - but then every agent is different, and the benefit of trying to scale by having a whole bunch of users connected to copies of the same model no longer works.
Reply
logainofhades

AI does nothing but cost me more money, be it via PC hardware prices, or increases on power bill due to datacenters. I have 0 use for it, and welcome the AI bubble bursting.
Reply
EzzyB

lazymangaka said:
It's an interesting technology that's deeply boring from an "average consumer" perspective. The vast majority of the things that these algorithms can do just isn't that useful for most people, at least in a direct "using the thing" kind of way. I get that we always need a "next big thing" to sell the new product, but this is a technology that would have been way better at just being implemented behind the scenes to make existing services better (voice assistants, image recognition, automatic captioning, image editing, to name a few) rather than being splashed around like the be-all and end-all of tech development.
I love the attitude, unfortunately that doesn't get investors particularly excited.

My example of a good use of AI would be something like a CAD program for architects. The AI part would deal with engineering things like loads and stresses as well as knowledge of building codes etc. That sounds quite useful.

But, what we're seeing is this huge push for artificial general intelligence which is just kind of pie-in-the-sky if you ask me.
Reply
USAFRet

EzzyB said:
My example of a good use of AI would be something like a CAD program for architects. The AI part would deal with engineering things like loads and stresses as well as knowledge of building codes etc. That sounds quite useful.
Plugins for all the major CAD applications already has that. No AI needed.
Reply

Show more comments