How the internet of things could fail us

Last week I wrote about how every company is now a technology company thanks to hardware costs declining, ubiquitous connectivity and cheap analytics enabled by the cloud. Most people call this the internet of things.

But focusing on what IoT is can mask what it enables. This week I’m going to talk about what society faces as every company becomes a tech company with a focus on the risks and downsides.

Basically, IoT lets us gather previously untracked information at scale, digitize it and derive insights from it. The scale is important because it is unprecedented. It’s not the size of the data that matter, but that it now covers many more aspects of our lives, from the humidity inside our home to how many steps we’ve walked in a day.

Want more analysis like this? Sign up for my weekly newsletter.

Not only do we have more information available, but it’s digitized. That puts it in less danger of disappearing and, most importantly, makes it easily searchable. So photos of you drinking in college can be matched with photos of you as you walk through a building. Digitizing information is not new but the scale of it is.

And because every company will have this power, it will no longer be limited to Google and Amazon. Your insurance company will do it. Fast food giants and your employer will too.

While we focus so much on the benefits of the internet of things, it’s important to realize that the same technology that can deliver predictive maintenance capabilities for a factory can also tell people if you’re home, if you’re pregnant or even predict how effective you will be as an employee.

That is, if things work together at all. Let’s take a look at the risks developing around IoT. Most people think of the risks in terms of spending a lot of money on a gadget that may not work well or for long, but there are larger factors to consider.

Instead of building the web, we’re building app stores: The current platforms we have for IoT are walled gardens, but a large amount of the value from IoT will derive from it being open so data can traverse platforms to track weather patterns, offer new business models and even predict the spread of disease.

Right now, everyone in tech (and some outside of tech) see a huge opportunity and want to own a slice of it by putting walls around it. Maybe it’s a proprietary data format. Maybe the pricing for the data is absurd. Either way, if we don’t get this right, the promises of the internet of things will never outweigh the downsides I explore below.

We are selling consumers cheap: When Jeff Hammerbacher, an early employee at Facebook, was quoted as saying “The best minds of my generation are thinking about how to make people click ads. That sucks,” all of social media was indicted. But given the data that companies can collect about consumers who use their connected products or shop in connected stores, the attempts to squeeze another dollar out of every shopper will only become more numerous and subtle. It may be something as subtle as offering a consumer directions that ensure they pass by a favorite store, or as obvious as telling you you’re out of milk and reordering another jug.

Some of these services will be helpful or beneficial to consumers. But consumers may not be aware of how much data they are giving up, how it might be used and how easily they can be manipulated. The combination of AI and the internet of things is a powerful tool that will be used to get consumers to spend…even if they can’t afford it.

Data becomes discriminatory: Does your car insurance plan change pricing based on your driving habits or demographics? Some insurers are already rolling out discounts for drivers who have connected devices in their cars and follow good habits like conservative braking. United Healthcare offers employers the option of giving their employees money toward their health savings accounts if they strap on an activity tracker and walk more steps.

For now, these programs are opt-in and incent customers with discounts, but it’s completely plausible that with greater transparency at the individual level, pricing becomes more personal and perhaps discriminatory. Even if your insurer doesn’t change your pricing plan, it’s possible that companies will use activity tracker data to make decisions about who gets promoted (that guy who performs well on 4 hours of sleep tonight over the woman who needs 9 hours).

Security breaches lead to physical world harms: CES was full of talk about IoT security and the Mirai botnet that took over IP cameras and other connected gear. However, the focus on poorly secured devices that have been on the market for years obscures the larger threat. Thanks to my connected devices, I can effectively stalk my family from afar by checking a humidity sensor to see if my husband has taken his morning shower yet, or getting a notification when my family heads out the door and walks by a connected video doorbell. I can see when people are in various rooms because I know what lights are on, and I could even use noise detection sensors to know when my husband and daughter are talking.

Taken as a whole this information gives a pretty clear sense of how we’re moving about and what time we do things throughout the day. I can even follow along with them in the car through my Tesla’s app or by using a device like Automatic. So while using devices to take down web sites is a significant threat, it’s really this cluster of information that proves a significant risk. If anyone were targeting me, they’d only have to access my phone or my husband’s phone (or account) to gain access to a wealth of information about my comings, goings and habits.

You also have to be aware of what your kids or spouse can see based on these data streams. Having an affair would be tough in a connected house.

Your privacy is defended by companies like Amazon: Related to my newfound ability to stalk my family is this idea that the data I use to do so is locked in servers at tech firms across the globe. Those companies may do a good job protecting that data or they may not. There are two levels of protection to think about here. The first is protecting users from malicious individuals and the other is protecting users against government overreach and attempts to access this data.

So as we connect our data and lives more deeply to the internet, it’s possible this whole experiment fails if we don’t think through the implications and guard against our worst tendencies as a society.