The just say no campaign was a good idea. It just should have been applied to social media and called tethics.
exformation 8 | Space capitalism + a dive into the infrastructure bill + more
Free at the point of consumption.
Delivered every week or so.
Friends, I apologize for the gap. I’ve been on deadline for some projects and had to buckle down. I am also trying to figure out the best format for this newsletter. One of the things I’ve heard is that there is a lot to consume. So I am going to spend some more time diving into topics. To do this, instead of every week, my updates will now come every other week. Much like the other work I have going on at CGO, I am trying an experiment to see what works. Let me know what you think.
Before I get into the news, I wanted to let you know I have a new op-ed over at the Regulatory Transparency Project with Adam Thierer on space capitalism. Here is the nutgraf: “Instead of dunking on billionaires, critics should take it as a chance to learn what has gone right and apply those lessons broadly. Smarter policy combined with American ingenuity is a recipe for success, both here on Earth and out in space.”
In 2005, NASA undertook a new way of contracting with the Commercial Orbital Transportation Services program, the benefits of which we are just now seeing come to fruition. Billionaires going to space is a byproduct of good policy. It says a lot that Bezos and Branson are able to buy a ticket, any ticket. This has never happened before and it is worth understanding. Let me know what you think about it.
Caden and I have been cranking on the podcast. This week we talked about the infrastructure bill and different ways to roll out broadband infrastructure. We are rounding out our first season and will be taking off a month to rethink the project. If you have suggestions, send me a note because I’ve got a lot of changes in mind.
News, notes & quotes
The big news is that an infrastructure deal has been worked out and there is a lot to go around. Here is the full 2700+ pages of what the Senate passed. The total bill is $1 trillion+.
Much as the reports had suggested, broadband is getting $65 billion. Just over $14 billion of that is going to extend the Emergency Broadband Benefit program, a smart move, I’d say. According to FCC chair Jessica Rosenworcel, more than 4 million people have signed up for the Emergency Broadband Benefit Program.
But instead of $50 per month, the new number is $30. What’s the right level of support? Is it $30 or $50 or maybe $70 for some families in need? It seems there will be some money in the budget to study this topic. Let’s hope the FCC actually conducts the experiments correctly.
A big part of the rest, $42.5 billion in all, will be going into a grant program that will be dispersed directly to the states and territories. I’m less enthusiastic about this one.
The process was exactly backward for the infrastructure bill. First, the size was negotiated, starting at $100 billion and then $40 billion, and then finally settling on $65 billion. And instead of conducting an auction, it will fund the states directly with a formula and instruct them to figure out the dispersals. There is sure to be high variability and some states are going to do much better than others in disbursing.
Instead, we should have followed the same steps that the FCC has been following in their funding of the unconnected. First, you map out all the regions that have broadband, set a threshold for service, and then calculate the general cost it would take to plug all those broadband gaps. The next step is to target those areas that don’t meet the standard through a support mechanism.
Beginning in 2018, the FCC began to support areas through a reverse auction. The support comes with a hook, providers to have skin in the game, and so the support grant typically covers around 20 percent of the cost to build the network. The most recent auction of this type closed in early 2020, the Rural Digital Opportunity Fund or RDOF. We are just now seeing that money head out the door and another round of RDOF is in the works.
On top of that, Covid relief money hit state budgets and many sepnt their dollars on broadband. Personally, I would like to know how well all of this other money does before pumping $40 billion+ into more. For a longer treatment of this idea, check out Gregory Rosston and Scott Wallsten in The Hill.
In the leadup to the passage, a WSJ article cited two studies “showing that expanding high-quality internet access could have significant economic benefits.” I did a little digging into the first study, from the Aspen Economic Strategy Group. It found that universal high-speed internet service would raise labor productivity by 1.1%, using the results of a survey.
Here’s the dirty little secret of broadband. It’s not the game-changer that many think it is. In the language of economics, it’s not a general-purpose technology or GPT, which would mean it would impact productivity. Electricity is a GPT and it has been shown time and again to impact productivity. Those trends are easy to find. As for broadband, it’s difficult to find a big impact on economic growth.
People don’t just buy Internet access, they typically adopt a bundle of technologies, a smartphone or a computer alongside an Internet connection. A meta-study of economic research, which is an empirical compilation of all studies to determine trends, helps to tease out the various impacts of the bundle. They found that broadband access doesn’t have an impact on economic growth. Rather, its adoption and use of cell phones and computers that drive new growth.
All the evidence suggests that broadband policy should contine to shift focus. The conversation is still dominated by the avilaibility gap and putting money into the ground even there isn’t good data on its size. There is a lot of money in that bill and very little of it is decicated to new pilot programs.
Extra: Austin Vernon has a great post on computers as GPTs.
Ookla released data on the speeds of Starlink. It is more variable than other services and seems to swing depending on the weather. Still, the median speeds are looking strong and much higher than their rivals in the satellite space.
Things I’m tracking:
Rep. Cathy McMorris Rodgers’ big tech accountability agenda now has some language. And the language comes in the form of 30 bills. Here is the extended list with links to the various texts.
Sen. Wicker just reintroduced the Setting an American Framework to Ensure Data Access, Transparency, and Accountability Act, or SAFE Act alongside Sen. Thune, Sen. Blackburn, and Sen. Fischer. Here is the text of the bill.
We still don’t know much:
NASA’s OIG report on spacesuits. They are late.
Papers & research
PrivacyCon happened last month, and, like in previous years, there was a lot of good research to read. Click on the agenda section on this page to find all the papers. Here are a couple that struck my fancy:
On the Privacy Risks of Algorithmic Fairness: “We show that fairness comes at the cost of privacy, and this cost is not distributed equally: the information leakage of fair models increases significantly on the unprivileged subgroups, which are the ones for whom we need fair learning. We show that the more biased the training data is, the higher the privacy cost of achieving fairness for the unprivileged subgroups will be…”
What Twitter Knows: Characterizing Ad Targeting Practices, User Perceptions,
and Ad Explanations Through Users’ Own Twitter Data: “We find many targeting mechanisms ignored by prior work — including advertiser-uploaded lists of specific users, lookalike audiences, and retargeting campaigns — are widely used on Twitter. Crucially, participants found these understudied practices among the most privacy invasive. Participants also found ad explanations designed for this study more useful, more comprehensible, and overall more preferable than Twitter’s current ad explanations. Our
findings underscore the benefits of data access, characterize unstudied facets of targeted advertising, and identify potential directions for improving transparency in targeted advertising.”
Reconsidering Privacy Choices: The Impact of Defaults, Reversibility, and Repetition: “We isolate three specific tenets of modern privacy regulation,
reduction in privacy consent opted-in by default, reversibility, and repeated consent, and explore their effects on individual behavior. We conduct an online experiment that presents participants with actual disclosure decisions that asks participants to link sensitive disclosures to personal information through the decision to ‘log-in.’ Expectedly, we find that active choice consent structure and a protective opt-out consent structure decrease log-ins significantly compared to
default opt-in. However, opposite the expectation that repeated exposure will lead to less susceptibility to choice defaults, we find that repeated exposure increases the effect of choice defaults, further entrenching their impact. For reversible consent decisions, we also find the surprising result that both explicit reversibility and explicit irreversibility increase the impact of protective defaults by up to 50%. We conclude that while opt-out defaults can drive more protective consumer behavior, in combination with reversibility and repeated exposure, they may the unintended effects of an over-reaction by consumers and lead to drastic reductions in consent provided. Our results extend the current privacy literature and have significant implications for consumers, firms, and policy makers.”
Models make meanings: A short little blog post on r-selected and K-selected species.
The just say no campaign was a good idea. It just should have been applied to social media and called tethics.
We don’t seem able to admit the obvious, that a sizeable part of the chronically online, the shouting class as Noah Smith calls them, display all the worse symptoms of the dark triad: psychopathy, narcissism, and manipulative behavior.
They are trolls and they pollute.
The common narrative has got it exactly backward. It isn’t that social media and platforms make us into monsters. Rather, a small group of digital deviants spoils the experience for everyone else. There’s a deep misunderstanding about the effects of social media and communication technologies, so I’m working, constantly, to update a bibliography on the topic. Here are some of the biggest lessons I’ve learned, which are always subject to revision.
First off, the Internet isn’t egalitarian. This should be painfully obvious. A small group of people tends to be the cause of a very large part of all content. Over a decade ago, it was found that “About 2 percent of those who start discussion threads [on forums] attract about 50 percent of the replies.” As study author Itai Himelboim, assistant professor in the UGA Grady College of Journalism and Mass Communication, lamented, “So although we have this wide range and diversity of sources, only a few of them are actually attracting attention.”
This power law turns up across the ecosystem. Around 10 percent of Twitter users create 80 percent of tweets. Reddit is built on a similar scale; the most prolific 0.1 percent of all users wrote 12 percent of all comments. Wikipedia too is largely the product of a small minority, 77 percent is written by 1 percent of users.
Online power laws create online power structures, and this is especially acute when it comes to political issues. Pew found that even in the middle of the 2020 election season, some 70% of U.S. social media users never or rarely posted about political and social issues. A small group, just 9 percent, is talking about politics all the time. If you talk about politics online, be honest with yourself. You aren’t a normie.
If you needed any more evidence to this fact, NPR disabled comments on their website after finding that more than half of all comments in a three-month period came from just 2,600 users. As even they admitted, NPR’s commenting system, which gets more expensive the more comments that are posted, and in some months has cost NPR twice what was budgeted, served a very, very small slice of its overall audience.
Online power structures lead individual users to believe in the majority illusion. As Lerman, Yan, and Wu explained, “In some cases, the structure of the underlying social network can dramatically skew an individual’s local observations, making a behavior appear far more common locally than it is globally. We trace the origins of this phenomenon, which we call ‘the majority illusion,’ to the friendship paradox in social networks. As a result of this paradox, a behavior that is globally rare may be systematically overrepresented in the local neighborhoods of many people, i.e., among their friends.”
The majority illusion is an extension of the friendship paradox. First observed by the sociologist Scott L. Feld in 1991, most people have fewer friends than their friends have, on average. That’s because almost everyone has that one friend that is extremely popular and known by many. That one person skews everything, pulling up the total average. Similarly, a single loud voice online can make it seem as though a relatively rare idea is actually more common than it is.
The majority illusion is double-edged. It can bring attention to much-needed political issues. While the number should probably be higher, 14% of Americans have changed their mind because of something they saw on social media. Indeed, around 80 percent of black Americas say social media help shed light on rarely discussed issues. It’s worth noting that the same share of whites says the opposite, that the sites distract from more important issues.
The majority illusion also explains the outsized influence of trolls. It only takes a few people to make online experiences horrible. Just one percent of Reddit accounts generate 74 percent of the attacks. As Noah Smith pointed out, “A study by the Anti-Defamation League found that two out of three anti-semitic tweets sent in 2015 were sent by just 1600 accounts.” As recent reports have reiterated, most of the vaccine misinformation comes from 12 people.
People do perceive online conversations, especially about politics, as being more hostile than in-person conversations. Could it be that nice people are less able to control themselves online? Research from Alexander Bor & Michael Bang Peterson suggests this isn’t the case. They also find that online environments don’t neccesarily bring in people predisposed for hostility. Hostile people talk about politics whenever they can. Nor is context collapse as severe as some might think. People misinterpret benign intentions as being hostile both online and offline in the same way.
Rather, "the public nature of online discussions exposes people to way more hostile attacks directed against strangers." Scaling up your network means scaling up the instances you might be trolled.
None of this offers comfort to those who receive constant trolling and harassment. But it should make us all think much harder about how we interact with people online and what we want to get out of the experience. We are in an age of digital adolescence. We need to mature and become more realistic about what’s actually happening online.
Part of that maturation requires better intentions and holding yourself to a high standard. People think Twitter and Reddit and Facebook are cesspools, but I have a very different experience with them. My inputs are varied, I actively filter out topics I don’t want, and I try to be intentional with these technologies. I try to be honest in my conversations, and admit when I was being too animated or aggressive as well. I use the Newsfeed Eradicator to keep me out of Facebook during work hours, and during work hours, I use Twitter and Reddit to find niche articles. On the weekends, I try to keep off of Twitter and instead focus on reading and hobbies.
Just say no has a really fun history in U.S. politics, but it’s striking that the idea has little purchase when it comes to social media. In its best reading, just say no was about empowerment. In tech, the idea of empowerment hardly registers, but it is there. You can see it in the work of Cal Newport, Nir Eyal, James Clear, and Marc Manson. It exists on the nosurf subreddit, ironically enough. The extensive use of ad blockers, Screen Time, do not disturb mode, and a number of other apps only confirms that people are learning to live with tech. In upcoming work, I will be marrying these threads with the idea of digital grace or technological atonement to help chart a new tech ethos. I think, rightly, it should be called tethics.
The Internet is an unsupervised playground. A lot of Grover Dill’s are out in the yard. You have to be the adult.