Ben's Website

Serious musings

Controlling Privacy

I've maintained that the real danger in digital privacy is when external actors know more about you than you realize you have revealed. This is particularly pernicious when they aggregate data and use statistics to conclude things you may not be aware of yourself, giving them the upper hand in any interaction.

Most people, Californians, Europeans, favor a digital privacy right. This set of laws instituting certain rights to one's digital information are an obvious, and obviously flawed exercise. With the FTC suing Kochava, a data broker, the alternative thought occurs to manipulate the market for this data.

To take one extreme future, assume everyone livestreams, and every interaction is labeled and able to be reasoned over. This rich data is itself accessible to everyone to analyze for myriad purposes. Much as we each occupy differing cultural niches today, so too will our analyses. Differing algorithmic tests for friendship, measures of political agreement, estimates of personal value, rankings of consumerism or environmentalism, indices of patriotism, the list of potential analyses goes on. Whereas today these aspects are decided by outward action, the increase in data and reasoning merely moves the decision threshold from outward action to internal disposition. The banality of this ocean of data quickly becomes apparent, as do the very limited circumstances when any of these analyses actually create value for society as a whole.

In the interpersonal aspect, having data and statistically-robust methods for regulating interpersonal relationships is inherently zero-sum and likely to further instrumentalize relationships. In the commercial view, more data just shifts purchase risk to some other domain.

"When the value estimate of my old washer went negative, I knew it was time to find another. I bought this one because I received that ad then, heard of Jerry's positive experience, read these reviews, and was finally convinced when I saw that their circular trade-in policy would make up for the last three weeks of the old washer's negative value." Even if each of these aspects can be influenced by various mechanisms, the basic decision, to replace an ailing washer, does not contain much opportunity for creating societal value, only for shifting that purchase from one supplier to another. And while the suppliers care about this, all that this abundance of data can do is make the purchasing decision more individualized. That is, instead of stickers with generic information about the average use of a washer and equally uninformed salespeople, both will be able to be more particular to your condition. This takes out much of the purchase risk, and would likely hasten the transition from private ownership to machine washing-as-a-service, purchasing a quality wash.

In this open extreme, decision making processes can be analyzed and corrected. Some people do their homework before making a purchase, limited by the information available, and the sophistication of their analysis. Outsourcing the analysis, or otherwise making them transferrable, can improve the quality of individual decisions. But as you increase the sophistication of the customer, you decrease the profitability of marketing. No longer can a supplier use flash sales or pricing gimiicks to drive purchases of less-than-ideal items. Likewise, loyalty rewards, coupons, and other forms of positive marketing will decrease in effectiveness. The point is that when you shift information within a closed system, you are merely redistributing the gains from that information among participants.

But for the impulsive consumer, excessive materialists, those occupied by vanity, hoarders, the availability of the best and worst analyses will give a finer point to their chosen lifestyle. Martha Stewart's home living may employ a suspect analysis on kitchen gadgets and housewares, but this can be tolerated and even chosen when someone is searching for a bit of feel-good distinction from their peers. Of course, if the economy tanks a different lifestyle could be chosen.

People already choose to follow Martha Stewart or fast-fashion or any of the multitude of influencers, the mere availability of data will not change their appetites. Nor, frankly, will wide data collection increase personal culpability or vulnerability. Invasions of privacy are that because they make certain private information public, the publication compromises your power of self-disclosure. But as cultural norms of privacy erode to permit the many possible uses of personal data, we will benefit if we invest in technologies that give us the same level of knowledge as the marketers. That is we can pro-actively neuter breaches in privacy by knowing ourselves better and understanding why we prefer what we do. When we know what they seek to, their data is effectively commoditized and holds little ability to manipulate.

For example, I'm in favor of video surveillance outside and inside my home. I am not in favor of Amazon or the government doing that. If I install cameras and recording equipment, if I choose to send that video data to a service for remote monitoring, I have control and knowledge over that process and can weigh the benefits of remote monitoring against the monetary or data privacy cost. If instead I install Amazon's cameras and pay them monthly for subsidized equipment and monitoring, I can achieve the same or better sense of safety, but I'll have no knowledge whether they are selling my preference in peanut butter.

The difference between DIY home monitoring and Amazon is that the former allows me to choose how I disclose my domestic information. Now if Amazon watches what I throw out to populate a grocery list, well there are times when that would be great. All Amazon has to do is to present users with an interface that allows them to run the same analyses as Amazon, to simply be open to sharing the insights gleaned from observing my life. They could give you opt-in control over the analyses, and if they proved their trustworthiness to your satisfaction you could share a wide variety of data with them. The point is that Amazon/etc. should not be seeking a lever over customers, an 'in' that allows them to manipulate people, but rather to empower users to live better lives.

Many people want personal assistants to help them manage the complexities of their lives, few want personal minders to decide for them and tell them what to do. My concern is that avoidance of the latter will prevent the former.