Testing my thesis statement (in the recent Project Snapshot), Paul Diduch pointed out to me that significant impacts are usually not hard to discern. That’s true, but there are exceptions. I believe orgregores are one of them. Here is first cut justification of my claim.
Two points up front:
- Discernment depends on the discerner. I’m claiming that the impacts are difficult for many/most of us to see, even though there may be people that are well aware of them.
- I focus on negative impacts. I’m assuming we can depend on Plato’s Theuth and Silicon Valley to highlight the positive.
Digital tech examples of "significant but hard-to-discern"
A first example is Tiktok serving up sex and drug videos to minors (WSJ, Sep 2021)—a significant impact. While it was known to the teens concerned, it wasn’t widely known. Or easily known: the WSJ had to create dozens of automated accounts and let them loose on the feed to gather evidence of what was going on.
The Journal reported that “analysis of the videos served to these accounts found that through its powerful algorithms, TikTok can quickly drive minors—among the biggest users of the app—into endless spools of content about sex and drugs.” It’s a feature of all these social media algorithms; by learning your most hidden interests and emotions they can drive users (of any age) deep into content rabbit holes that are heavily dominated by videos about a specific topic or theme. YouTube has had the same problem.
For a view on the risk TikTok poses to the U.S. as a country, check out Scott Galloway on a recent Realignment podcast, from around timecode 33:41 (the question is posed around 32:30; he gets into geopolitics around 35:05).
Second, the Journal has also reported that Facebook knows Instagram is toxic for teen girls (WSJ, Sep 2021). The paper alleges that the company’s “own in-depth research shows a significant teen mental-health issue that Facebook plays down in public.” Again: a significant impact that some people knew about, but not the public at large. There are more examples in digital tech but I haven’t inventoried them. These examples suggest that they require resources to uncover.
Perhaps I can make the claim more plausible to you by offering instances of significant but hard-to-discern impacts in non-tech settings. I’ll then return to speculating about how tech orgregores influence us, and why it’s hard to see.
Non-tech examples
Public health and the environment offer many examples. There are many substances that we now know to be toxic that people didn’t worry about in the past, like radon in the basement, lead in gasoline, and asbestos in heaters and building materials.
One must beware of anachronism: things that are obviously and patently dangerous today weren’t always perceived that way in the past. At the moment, few people and few experts worry about radiation from cellphones even though there are a doomsayers. Perhaps the Cassandras will turn out to be right—or perhaps not. It’s a too soon to tell.
Human-generated CO2 was already having a significant impact on climate fifty years ago—one could see it in the temperature data—but most people didn’t rate it as an important issue since it wasn’t impacting their daily lives through heat waves, droughts, and storms, as it increasingly does today. Even now, it takes sophisticated science to figure out the extent to which a particular extreme weather event can be ascribed to climate change.
Turning to politics: I believe the pharma and healthcare industries have a significant impact on the cost and delivery of medical care in the US, but I also believe that this influence is obscure to most people. It’s not invisible, of course. Opensecrets tracks who spends what lobbying the government. In 2021, three of the top six spenders were healthcare interests:
- US Chamber of Commerce ($66,410,000)
- National Assn of Realtors ($44,004,025)
- Pharmaceutical Research & Manufacturers of America ($30,406,000)
- Business Roundtable ($29,120,000)
- American Hospital Assn ($25,215,934)
- Blue Cross/Blue Shield ($25,176,385)
The military-industrial complex also has a significant social impact, largely through absorbing public funds that aren’t spent on other things. One can argue that Europe’s idyllic half-century of social democracy was funded by under-investing in its military since they felt safe under Uncle Sam’s umbrella. Here’s another excerpt from Opensecrets’ 2021 data showing the heft of defense contractors on Capitol Hill:
12. Raytheon Technologies ($15,390,000)
13. National Assn of Manufacturers ($15,300,000)
14. Lockheed Martin ($14,401,911)
15. NCTA The Internet & Television Assn ($14,010,000)
16. AARP ($13,680,000)
17. Boeing Co ($13,450,000)
There are of course also positive impacts that are hard to see. On a different scale but sticking with the health theme, scientists and many members of the public understand that the gut microbiome is important for our health. It’s always been like that, but we’ve only recently come to understand it.
Back to digital tech
Algorithms
It’s hard to discern the workings of the algorithms that decide what content to show us. That’s also true of a TV or paper newsroom, a Hollywood studio, or a boardroom, but they’re easier to imagine since people are making the decisions. Algorithms are completely opaque to the public and largely opaque to their operators. Algorithmic filtering is everywhere; it’s the water we swim in nowadays. That makes it even harder to discern. Because their workings are hard to discern, cause/effect is obscure and it’s hard to assess their impact. I think it’s significant, though.
Algorithms are filtering the content we see. Some voices are amplified and muted. Depending on the issue and one’s political views, the intervention may be too much or not enough. Either way, only a few activists and not the general public are worried about algorithms. Algorithms’ political impact may be decreasing as Facebook and Google back away from news and political content, which was always a relatively small part of their mix, but the impact persists in shaping entertainment. Content creators are constantly complaining about the quirks and vagaries of the algorithms. Algorithms are now helping shape what counts as entertainment, and what doesn’t—a significant impact.
Market concentration reduces the diversity of algorithms and just a few algorithms are doing most of the filtering. In the US, best I can tell, content filtering is dominated by Alphabet (Google, YouTube) and Meta (FB, Instagram). So not only are algorithms deciding the broad contours of culture’s content, but only a few algorithms are doing the work.
Shaping desire
The companies are in the business of influencing desires and actions because they sell advertising. On the one hand, there’s nothing new in marketing. On the other, though, the amount of surveillance and quality of information on what users are seeing and how they’re reacting to it is unprecedented. Note that the surveillance is hard for most of us to see; few people use ad blockers, and even fewer click on the “how many trackers blocked on this page” button. Similarly, while some people may be aware of the range of information that (say) Google is connecting about them, only industry insiders understand how such information is aggregated and amplified to create personal profiles.
Shaping values
Institutions have a significant impact on us both individually and collectively, but we’re largely insensible to how they influence what we can think (cf. Mary Douglas, How Institutions Think). Technology is an institution and so influences what what’s thinkable and how it’s thought. Technology isn’t value neutral. Widely deployed technology imposes its values widely.
Intangibility
It’s not just the impact that’s hard to discern. The mechanisms of digital technology (i.e., the causes and means of the impacts) are largely intangible and invisible. Digital tech has a high bit/atom ratio, and one can’t figure out what software is doing the way one can see the gears going round in a flour mill or factory. Internal combustion technology shaped our society, but one can at least see the engines and highways (if not the lead in gasoline). It’s much harder to see the algorithms and data networks.
The service economy is less tangible than the manufacturing one, but at least we can observe many moving parts of industries like fast food and healthcare. Once we get to the information economy, the opacity increases. Even before digital tech, some industries were less tangible than others, notable finance which is pure information (especially after Bretton Woods). Meta-instruments like futures and options have had a significant impact on markets for centuries (cf. Investopedia’s history of debt securitzation). It takes melt-downs like the 2007 crash built on the securitization of debt for the public to realize how much is going on below the surface.
Scale
Last, we get back to scale. The biggest ad networks seem to be Google and Facebook—concentration again, now in how the internet economy is funded. The online advertising business is booming. Spending on pure-play internet ads as a percentage of total ad spend grew from 16% in 2012 to 49% in 2020, and is forecast to be 60% in 2024 (VisualCapitalist). Pure-play internet ad spending was $109 bn in 2020; in that year, Google’s revenue from advertising (which is not just pure-play, I guess) was $147 bn (Statista). In 2020, Meta did $84 bn ad revenue (Statista).
The sheer scale of digital technology makes it hard to grasp. For example, people watch 1 billion hours of video per day on YouTube. I make that about 40 million people at any moment, all watching clips served up by the same algorithm. It’s hard to grasp numbers that big—here’s an attempt to illustrate it. Imagine Hangar 1 at Moffett Field filled with fluttering butterflies; that’s about 40 million. Or a school of anchovies the same volume. Or the hangar packed floor to ceiling with large frozen turkeys. Each butterfly, anchovy, or turkey represents a user watching a YouTube video.
Two tangents to close:
- A la McLuhan, algorithms are now the medium through which many of us get much of our information. What is the message that we’re getting from the medium?
- It’s been a while since I read Technopoly, but Postman’s arguments that grew out of a TV era are probably supercharged by post-1992 technology.
No comments:
Post a Comment