I survived CES 2017 to tell this story. From the very first day of press conferences and CES Unveiled, it was obvious connected-anytime-anywhere was what we were going to see at the show. It soon became clear that, if we played the drinking game for every time AI was mentioned by a vendor, we would not be sober for very long. Press conference after press conference, pitch after pitch, Artificial Intelligence was mentioned as a key trend for 2017 and one vendors were working on. Although in most cases, there was nothing concrete to see linked to the product they were announcing.
Everything Is Connected Even When It Should Not
A quick Google search pins the origins of the Internet of Things to Peter T. Lewis back in 1985 when he gave a speech at a FCC-supported conference where he defined IoT as “The integration of people, processes and technology with connectable devices and sensors to enable remote monitoring, status, manipulation and evaluation of trends of such devices.”
Over the past two to three years, IoT has started to materialize as more and more devices were connected to the internet. How this is developing might be different than how vendors intended it a few years ago. If we rewind a little, after smartphones, many thought wearables were going to be the most pervasive devices, making humans the most connected ‘things’ of the Internet of Things. As it turned out, wearables penetration is ramping up very slowly and it seems the focus of the vendors and the attention of the consumers have both shifted to connecting our homes.
What gets connected, however, is questionable. Not everything that is connected should be. Sometimes, even if a device is connected with good reason, the value of that connection is not immediately clear to the customer. In 2013, the HAPIFork took CES by storm. The connected fork let users know when they ate too much or too fast and was one of the most talked about device of the show. In 2017, one of the most talked about devices at CES was the L’Oréal Smart Brush. Developed with Withings, the Kérastase Hair Coach uses sensors and a microphone to count hair strokes and listen for hair breakage. Aside from the brush, I saw connected showers promising to keep your water warm (and save you water and money as well), windows that can sense when the air is getting too stuffy and open on their own, smart locks, voice activated garbage cans and so much more. There was so much connected “stuff” that I started to mentally file gadgets into three categories — Tech for the sake of tech, tech for the sake of lazy, and tech for the sake of humanity.
Sadly, I saw more gadgets and solutions falling into the first two categories than into the last. There were many products searching for a problem to solve and many that offered to replicate something we already do today with just less effort on our part. What was interesting, however, was the common underlying selling point was the smartness, not the connectivity.
IoT Is No Longer Cool for Consumers
I came away from CES with the clear impression that, although we are still talking about the Internet of Things, vendors, and more importantly PR gurus, have moved on from the connectivity part to the brain part of the devices.
While enterprise is still very much talking about IoT as it looks to empower, manage and, most of all, monetize all these devices, it was as if the term was tired when it came to consumers. It could also be for consumers, being connected is a given nowadays and it is the value of connectivity that needs to be highlighted in order to drive a premium. Unfortunately, not everything that is connected is necessarily smart and not everything that is smart is necessarily intelligent. As I talked to people, I noticed the line between these three concepts was very blurred and, in most cases, the blurring was quite intentional.
For me, a smart device is not only a device that is connected to the internet and/or other devices but one that can interact with other devices and enable some degree of autonomy. In order to be smart, a device does not necessarily need Artificial Intelligence, at least if you think of AI in terms of a device that mimics “cognitive” functions that humans associate with other human brain such as “learning” and “problem solving. This does not make the device less smart or less useful. To give you an example, let’s look at the iPhone 7 Plus camera experience. The iPhone 7 Plus takes great pictures thanks to the two lenses, the software, the sensors, and the processing power. AI or, more specifically, Machine Learning, only comes in to deliver the portrait effect by recognizing where the subject of the picture ends and where the background begins. Think about maps as another example. The Estimated Time of Arrival (ETA) we are given when we set a destination comes from a series of data from average speeds, actual travel time, traffic prediction, speed limits, and historical averages. These are all combined to get a projection of your ETA between two points. AI is what makes it possible for my phone to analyze my commute pattern and, as I connect to the car at a given time of day, offer me the ETA to the most likely location — my daughter’s school in the morning, the office after that or the Karate Dojo in the afternoon.
The Risk of AI-Washing
Internet of Things, Artificial Intelligence, Machine Learning are all trends that will develop over time, not over-night. While the temptation of keeping things fresh might get vendors to chase the next buzzwords, I think it is very risky to do so. Talking prematurely about features and capabilities might just bore consumers sooner rather than excite them. Talk about what your products can deliver. Better yet, showing what they deliver is more effective than labelling features with sexy buzzwords. Don’t tell your users your device uses AI, show them what your device can actually do and, if it looks like a bit of magic, let them think this is how you do it. Sometimes, talking too much about what is under the hood might raise more questions than you have answers for — especially when it comes to security and privacy – something we should all concern ourselves with as everything gets connected and smart around us.
4 thoughts on “Move Over IoT, AI is the New Hot Acronym”
I have a several years difference of opinion with Ben Bajarin over the definition of things and the extent to which consumers should be “allowed” to make technical and scientific definitions.
At the peril of being a pompous ass (which is irrelevant anyway), leaving these things to consumers is akin to practicing medicine without a license.
Marketing is another story…
I agree with the old adage that engineers would try to market Sushi as cold, dead, fish. (But that’s only because they’re honest). There should be, well, more “palatable ways” to promote things. But NOT at the expense of mis-defining them or mis-classifying them. Sure it may seem semantic, but it is substantive.
Your closing paragraph says it all, and I wholeheartedly agree. I’ll summarize my agreement in two quotes:
“IoT Is No Longer Cool for Consumers” – C. Milanese
“You think that I don’t even mean
A single word I say
It’s only words, and words are all I have
To take your cash away” – klahanas modified Bee Gees Lyric
“IoT Is No Longer Cool for Consumers”
When was it?
Even at the most computer-like level… iBrother got his 11yo son an xbox for Xmas. it took over a day and several tens of $CAD in bandwidth overage to get it up and running because the thing downloads hundreds of megabytes of OS updates and games updates/content (on top of the disc). He really wishes this T were not so Io. Online gaming is fine. Hours of wait and spoiled gratification because the thing can’t stand on its own legs, isn’t.
I am *not* eager to have all my gizmos behave this way.