CES 2024: AI pillows and toothbrushes – is all of it getting a bit foolish?

“Come get your AI pillow – stop snoring tonight!”

I’m strolling round CES – the tech business’s annual showcase of all its newest devices – in a little bit of a daze, till this pitch grabs my consideration.

What on earth is an AI pillow?

Motion Sleep, a South Korean firm, has a big area in one of many important exhibition halls at CES. Intrigued, I wander in.

First I’m provided a couple of stats concerning the penalties of dangerous sleep. One signal factors to the variety of accidents brought on by drowsy driving. Another goes by means of the well being penalties of sleep apnoea.

The answer the corporate has landed on is a pillow that detects loud night breathing. It then pumps air into completely different compartments of a pillow, which gently lifts the top, making the offending snorer roll over and – in idea – assuaging loud night breathing.

This is a fairly typical CES product to this point. There are hundreds of those sorts of innovations which will or could not take off.

But this pillow is completely different, we’re informed. This pillow incorporates AI.

“With the AI, it can be trained to know what you sound like specifically when you snore,” a consultant insists.

“That way it can differentiate between you snoring and the TV or cars outside.”

And the pillow is much from the one gadget to put declare to the particular energy that, we are supposed to consider, AI confers. At Samsung’s exhibit, for instance, a whole part was dedicated to AI-capable family gadgets.

I’m proven an AI vacuum cleaner – that appears very very similar to a traditional vacuum cleaner – with one small distinction.

An “AI” operate mode apparently permits the vacuum to evaluate sorts of surfaces. It can then apply completely different ranges of suction accordingly.

There’s an AI washer that may purportedly detect several types of material too.

“AI Wash uses sensors to sense the laundry’s weight and level of soiling, and optimises the amount of water, detergent and rinsing time, using machine learning,” a Samsung press launch says.

I’m proven the washer by a Samsung consultant. I’m nonetheless barely puzzled about how that is truly utilizing AI.

“This will learn your clothes,” they are saying. I stay confused.

Elsewhere on the present, there’s an AI mirror and even an AI toothbrush. No product is just too boring or humdrum, it appears, to flee an AI makeover.

The rationalization, maybe, is that each one of those corporations are dealing with strain from buyers and shareholders to have some type of AI providing, as a result of it attracts consideration and funding.

OpenAI’s extremely profitable launch of ChatGPT is why everyone seems to be speaking about AI – and the potential it has.

This is a big language mannequin (LLM) that makes use of a kind of machine studying to provide detailed and human-like solutions to questions.

But an AI toothbrush or vacuum is a really good distance away from ChatGPT.

And that takes us neatly to one of many main issues with AI extra typically. It has no universally accepted definition.

“AI suffers from an unrelenting, incurable case of vagueness,” Eric Siegel, a machine studying professional, informed me over the summer season.

That lack of a definition signifies that all issues AI have been caught up in a blistering yr of hype.

That’s the case with merchandise that already contained AI with out a lot fanfare earlier than. Now their AI functionality, nonetheless obscure, is hammed up.

But there’s a drawback with this craze for all issues AI: Companies claiming AI functionality when actually their merchandise do not truly use machine studying.

The Federal Trade Commission within the US has put out advisory notes aimed toward corporations stretching the definition of AI.

“Does the product actually use AI at all? If you think you can get away with baseless claims that your product is AI-enabled, think again,” a observe from the FTC printed in February final yr says.

Yet in that very same observe, the FTC accepts that AI is “an ambiguous term with many possible definitions”.

That’s clearly an issue for shoppers – but it surely’s additionally an issue for journalists. For years now I’ve coated corporations that declare to be utilizing AI of their merchandise. Often they supply no proof to again this assertion up – typically saying the know-how is proprietary.

It’s very laborious then to know what the engine seems like when you’ll be able to’t look underneath the bonnet.

Some corporations are already conscious that the usage of the time period AI has turn into counter-productive. One product that has had rave evaluations this yr at CES is named R1, made by Rabbit. The telephone like gadget makes use of a type of generative AI – and permits customers to bypass apps, and easily ask for issues to be executed. Like reserving a flight or taxi.

But in Rabbit’s pitch, AI is barely talked about. Instead the corporate talks about basis fashions, and even a brand new time period: Large Action Models.

Back at Samsung, I come throughout a product that clearly does use generative AI: A fridge that analyses the meals in it and might recommend recipes. At final one thing that seems to be fairly clearly AI.

But then I’m hit with one other query. Do I want my fridge to present me recipes? I’ve by no means felt pissed off concerning the lack of culinary recommendation given to me by my fridge.

So for now, I’m going to stay to my non-AI succesful pillow, my non-AI succesful washer and my non-AI vacuum cleaner. And, I believe, the probabilities are you’ll too.