Ad-tech and fatigue

The world is not getting crazier or more outrageous. The world has always been deeply strange, and it has always been full of injustice, horror, and loss. Communications technology has made it easier to remain aware of this, but more importantly, ad-tech has incentivized it.

Causing panic for personal gain is not a new technique. However, it used to work in two ways:

  1. A state of panic lowers people’s guard, making them willing to accept less rigorous arguments, and thus making them easier to manipulate in particular ways. Arguments only need to sound convincing.
  2. You can manipulate people by manufacturing a threat and then presenting your product as a solution.

Ad-tech changed this.

Panic (or, more precisely, autonomic arousal) creates a state of shallow vigilance: we are constantly scanning our environment for evidence of threats, but energy is diverted from the neocortex to the skeletal muscles to prepare for fight or flight. Shallow vigilance means increased engagement metrics: we refresh the page in order to check if the danger has passed, or if we need to do something other than freeze in place; we refresh the page again because our starved brain has already forgotten what we read ten seconds ago.

Ad-tech-driven renvenue models incentivize maximization of impressions (which are easily measured), not effectiveness (which has never been reliably measured), and so they optimize for engagement. This is not completely irrational: the mere exposure effect produces some benefit for advertisers, and does not depend upon reason, autobiographical or procedural memory, or any of the other facilities that are heavily impacted by panic.

Panic for engagement’s sake doesn’t need to be packaged with hope — we aren’t selling the remedy, after all. Panic for engagement’s sake doesn’t need to be temporary — we aren’t trying to foment an ill-considered riot and then turn former rioters into careful statesmen. It’s a new situation, because somebody with the systematic power to keep everybody panicing all the time also has the incentive to do so — and no incentive to suggest a way out.

Of course, constant low-grade panic has a body count. Arousal doesn’t merely redirect resources away from the neocortex — resources are redirected away from the immune system (so we get sick easier and remain sick longer) and the digestive system (so we are less efficient at extracting nutrients from food, experience intestinal discomfort, abuse our anuses with periodic torrents of diarhhea and our gullets with periodic torrents of vomit). Arousal forces the heart to work harder (increasing the likelihood of heart attacks & strokes), optimizes motor control for sudden forceful movements (making us jittery and unable to execute fine motor control), tenses up the muscles (increasing jitteriness and producing muscle fatigue), and generally wastes resources. Normal endocrine regulation of mood is put on the back burner — since this is an emergency, damn it — and so our normal protections against depression, mania, neurosis, and obsession are hamstrung.

On the other end, ad-tech acts as a broker between parties gambling on the value of ad impressions — both of whom, acting on models of advertising based on television and newspapers, assume that their advertising is operating on something less weak than mere exposure and therefore systematically overvalue it.

Mere exposure is important at scale: if you are Coca Cola, Pepsi, or McDonalds, your name is said so often that merely saying your name creates cravings; however, only such big players can afford to enter that game. If you aren’t McDonalds, you can’t afford to depend upon mere exposure: you need to reason with your customer, and for that, you need a customer capable of reasoning.

Even with both parties systematically overvaluing ad impressions, our impressions are still valued at a fraction of one one-hundredth of a cent. The broker can make plenty of money at this, by brokering billions of such deals per second. The host makes next to nothing, and with the price of impressions in free fall, must host more and more ads. The buyer of the ad spot gets next to no benefit from each ad impression in terms of actual sales, and wastes money buying more and more ad spots.

(Into this landscape walks the ad-blocker, which provides some tough love. The user has been screwed on this — footing the bill on cycles, electricity use, time, and network use, to a tune of several times what anybody else has made per ad impression — so he blocks the ads totally, saving himself money and pain while throwing the entire ad-tech ecosystem out of the casino.)

Of course, the news sites and social media systems are still optimized for ad-tech. They still optimize for engagement — even when they host no ads — because of a cargo cult approach to design that doesn’t consider their real goals. So the user has saved a couple cents on his electric bill, but he’s still getting sick.

None of this is “technology”. It’s not inherent to computers.

We made a mistake by accepting advertising as a revenue model on the internet. We made another mistake by forgetting that other models existed, once prices started crashing. Now, the people who are making the decisions are too sick and tired to see a way out because they are hamstrung by the system they created, and the people who need out the most can’t afford to take a break.

Computers and networked communication can be wonderful for the world, if we keep the world in mind when designing how we interact with them. If we keep our paychecks in mind, we lose both our paychecks and the world.