Having heard about the terrible privacy policies of car manufacturers and Ford’s new patent to automate vehicle repossession, I was finally able to formulate some thoughts I’ve had brewing for a while about ownership in this world of ubiquitous computing.
Before the industrial revolution, humans could essentially understand what a thing did by looking at it, and ownership was about possession of the object. Now we don’t understand our tools, and so possession does not imply control.
I’m not saying everyone who uses a computer should fully understand the hardware and software, because that’s unreasonable. But a company with a product that contains software can bundle arbitrary behavior into their product, which the user does not have control over. A pre-industrialized analog would be buying a house and seeing that there’s a hole in the roof. You can observe by inspection (either immediately or when it rains) that the hole exists, and then you can patch it yourself or get someone you trust to fix it. This generally keeps people from buying and selling houses with holes in the roof, because it’s clear by inspection what it’s doing.
But it’s different with software. The average consumer can’t inspect their product’s behavior. What’s more, the behavior can be arbitrarily complex, and it gets worse when there’s any reason for there to be a network connection. Whether the excuse is feature updates, security patches, or usage telemetry, the door is now open for them to add any behavior they want at any time. Whatever the original reason was, now there’s a continued relationship between the object and the seller.
This is most obvious with the Ford repossession thing, but it’s also pretty obvious with Amazon Echo/Google Home/etc. You own the speaker and the microphone, but the actual behavior of the object is entirely out of your control. Even if you fully understood it when you bought it, you don’t have any guarantees that things haven’t changed the next day.
All of this complexity in the objects we purchase is reflected in the corresponding terms, conditions, and privacy policy (T&C/PP). When you buy a (non-smart) sewing machine, you didn’t receive any T&C/PP because the machine (hopefully) just does what it’s supposed to do. You could harm yourself by using it in an unintended way, so maybe there’s a warning on the bottom about the needle being sharp in California, but that’s also obvious by mechanical inspection. But with “smart” products, so much explaining is necessary because the behavior is not obvious by inspection.
Putting all this together, having OTA updates and T&C/PP for a product results in there being an ongoing contract between the buyer and the seller, one that has implications for how the object is used by the buyer in the future. But this contract is entirely under control of the seller, and can often be changed at any time in the future. The seller has an army of lawyers to balloon the contract to unspeakable proportions, meaning that the buyer usually does not fully understand the contract. And there are no alternative products without this garbage, because it is in every seller’s interest to include this imbalanced power relation in the product. In this way, an important aspect of the meaning of “ownership” is lost, handed over to the creator of the product or service. That aspect is the control of the behavior of the object.
We’ve dealt with this before
This isn’t the first time we’ve created products that their users don’t understand. This also happened with food, and so we created the FDA. 👀