GenAI is the Internet of Things 2.0

Image: Badly edited by author

The Internet of Things has been a bit of a failed product strategy.

While it made sense at a high level—taking devices, products and appliances and connecting them to the Internet so they could offer new features or utility—there have been very few cases where anyone figured out how this connection could lead to something useful. Apparently, utility was low down on the list of priorities; high on that list was having something that companies and execs could show off at consumer expos or tech demos that made them appear relevant.

As time passed, it became clear that the IoT era was just another daft trend, resulting in all sorts of smart homes, devices and appliances that were, truthfully, dumb as hell.

Some notable highlights; smart locks that could be hacked (or fail to open due to software issues), smart fridges that could let you see inside them without opening the door, $700 juice makers that didn’t make juice, smart toasters with “7 shades of toasting,” smart cars that distract drivers with screens and alerts, smart speakers like Alexa that definitely don’t listen to everything that goes on around them, smart kettles that let you boil water from a different room, smart mirrors that told you the weather, and $8000 “intelligent toilets” that provide LED mood lighting and music, and the ability to control the temperature of the seat itself. There will be countless others — feel free to drop the best (worst) ones in the comments!

Like many other tech trends, those who had told us this future was inevitable had backed themselves into a corner. Now, they had a bad dose of sunk cost fallacy. They told us that things around us would talk to each other and be controlled by various apps, voice commands and random claps. They added Internet capabilities, cameras, microphones and screens to everything. They’d set off on this path, sure of its success — or at least sure enough we were going to eat it up — and there was no going back. Yet… most of us didn’t care. It turns out that consumers never bought into the smart revolution. It’s thought over 50% of customers don’t even connect their devices to the Internet, let alone use the features.

It actually put many of us off buying these products. The reality is that IoT devices have made our lives more difficult — truly the hallmark of modern technology. Adding tech elements made devices more likely to break. It made setup costs higher and more tedious. Running these products on clunky operating software meant they could break, be buggy or fall prey to viruses. We have yet more things that need to be updated before they can work from the box. The new “capabilities” meant many of our devices sucked up more energy to run. Adding Internet connectivity and the various cameras, microphones and screens, opened these devices up to privacy issues and hacking. And the pièce de résistance — our devices and products became tools for data harvesting.

The tl;dr here is that the IoT strategy made most products worse, ruined user experience and had a detrimental effect on safety, all in return for pointless features that many didn’t even bother to use.

All of the above sounds very familiar to the current ‘Next Big Thing’, doesn’t it?

GenAI has all the hallmarks of being the Internet of Things 2.0.

Despite VCs pouring billions of dollars into the industry, despite every company, product and service forcing it upon us, despite our tech overlords insisting over and over that it’s the future, generative AI has yet to make the majority of products or services it’s stuffed into any better or more useful.

In many cases, it has made them worse, made them harder to use, and increased their costs or subscription fees. The similarities to the IoT era also exist in the agenda surrounding GenAI; VCs have pumped the industry full of cash, companies have pivoted entire product strategies, and the tech industry is now essentially propped up on the hope that AI pays off big time. It means there’s no going back — they have to keep stuffing AI into everything, whether it makes sense or not, whether they want to or not, because they need to keep the line going up. Once again, utility is low on the list of priorities.

We’ve already seen a few standalone AI devices enter the market (the Humane AIPin, the Rabbit r1 and the Friend pendant) and all of them sucked. But that won’t deter anybody, will it? The next play seems to be to repeat the IoT playbook and stuff GenAI capabilities into all the devices and services around us, opening them up to the same issues — privacy and safety concerns, buggy software, over-complicated user experiences and more.

As I said, technology seems doomed to repeat its past mistakes.