Technology and the future

Eli Dourado has a nice roundup.

Both Moderna and BioNTech have personalized vaccine candidates targeting cancer. Although called a “cancer vaccine,” the treatment is only administered once the subject has cancer—it isn’t preventative. The companies use an algorithm to analyze the genetic sequences of the tumor and the patient’s healthy cells and predict which molecules could be used to generate a strong immune response against the cancer.

He is optimistic about the Hyperloop. Peter Diamindis is, also.

Pointer from Tyler Cowen. The tone is optimistic. Years ago, I noted that the print publication that gives me the most optimism is Technology Review. The one that gives me the most pessimism is Regulation.

We can compare Eli’s post to what I wrote in 2005.

I wrote,

Is the new trend rate of productivity growth 3 percent or higher?

Obviously, it wasn’t. I was thinking about labor productivity, which is not the same as TFP, which Eli wants to look at. All productivity measures are dubious, in my opinion.

I wrote that cognitive neuroscience was the field of the future. I think I was wrong. It gets no mention in Eli’s survey.

I quoted the U.S. Department of Energy to the effect that by 2015 solar power would be competitive with traditional sources of energy. Eli says

The 2010s were the wind and solar decade. We observed stunning declines in the cost of both

I wrote about optimism regarding cancer therapy. We have seen some improvement, but “the emperor of all maladies” is still a formidable killer.

Finally, I wrote that mainstream media were losing attention as well as credibility. Actually, they figured out how to hang on to attention by dialing up the outrage and clickbait. They decided not to bother with credibility.

If I were a billionaire, instead of wasting money on non-profits, I would buy some land and try to build a start-up city. Build roads that work well with autonomous vehicles, and only allow autonomous vehicles–visitors have to leave their ordinary cars in a parking lot outside the city. Design and build an electric grid optimized for today’s technology. Make 5G available everywhere. Implement a set of protocols that allow all sorts of drones to operate without colliding with one another or with humans.

The key to a start-up city might be making it easy for people to interact with people from other cities. This could mean rapid transit in and out, or it could mean the use of augmented reality.

Perhaps in the future cities that have major impediments to livability will be at a disadvantage. It may prove easier to give people in other cities the amenities that they appreciate about NY or SF than it is for NY or SF to escape their downward spirals. Boston, Chicago, and Minneapolis will have a harder time convincing people to live with winter.

That is not a sure-fire prediction–it may prove no more prescient than my 2005 blog post–but I offer it as a possible future scenario.

If I were in charge of content moderation

At a social media company, I would start with clear terms of service. See yesterday’s post.

I would have two types of user registration. One is “true identity” and the other is “anonymous.” I would require each anonymous user to pay a $10 non-refundable registration fee, because I want to limit the number of anonymous accounts. Both types of registration would be required to abide by the terms of service. Content by anonymous users would be labeled as such.

Human moderators would be the heart of the system. To economize on these resources:

All content would be run through an AI system that would assign a score to the content, with 0 for “apparently totally safe” up to 100 for “apparently in clear violation of the terms of service.” When the score is above a certain level, say 50, the content would be referred to a human moderator for auditing.

The scoring system would be updated regularly based on trends in the results of human audits. But I would not trust the AI system entirely. I would add random samples of the following types:

–regular random samples of content uploaded by users who have many followers.

–a tiny but purely random sample of all content uploaded. The point of this sample would be to make sure that humans agree when the AI system assigns low scores. If a human audit sees a randomly chosen item as being close to a violation of terms of service, this is a sign that the AI algorithm needs to be improved. Most users would never have any content picked up by the purely random sample.

All samples would be sent to the human employees for moderation.

When humans find content that violates terms of service or “comes close,” this would trigger various actions.

–If it definitely violates the terms of service, the user would be asked to remove the content within 24 hours. The user would be informed of the specific way in which the content violates the terms of service.

–The user would be put on a “watch list” that would involve increasing the rate at which that user’s content is sampled for auditing by humans. One way to do this is to lower the threshold for triggering an audit. Suppose that for ordinary content the trigger for an audit is a score of 50 or higher. For someone on the watch list, the content might be audited if the score is 20 or higher.

–The user’s network of followed/followers also would have their content sampled at an increased rate for auditing by humans. Maybe if the score is 40 or higher.

Users who repeatedly violate terms of service would be given a warning. Those who fail to heed a warning would have their accounts terminated.

Tyrone on social media censorship

Supposedly, Tyler Cowen wrote this about Twitter’s decision to ban President Trump:

I am fine with their decision. Furthermore I think they made it at exactly the right moment.

But Tyler has an evil twin, Tyrone, who occasionally takes over his blog posts.

Here are my thoughts on “terms of service” and their enforcement.

1. In my case, I dislike comments that include personal attacks, especially when they are personal against other commenters or personal against me. Criticize a person’s ideas all you want, but avoid name-calling and insults. I do not have the capability to enforce this all of the time, so sometimes inappropriate comments do get through. But apart from personal attacks, I have not found myself wanting to delete comments.

2. Terms of service should be clear, not vague. You should try to make it as clear as possible how to distinguish what is ok from what is not. Examples are helpful.

3. When you delete someone’s content, you should have a good answer to the question, “You deleted X, but why did you not delete Y?” Ideally, you should be able to show that your terms of service allow Y but not X.

4. Banning an act is serious, but banning the actor is grave. Banning a person should be a very last resort. It seems right to ban someone only after they have been found multiple times to have posted content that is against your policy, you have warned them that their abuse is excessive and could result in a personal ban, and they continue their abuse.

5. When you ban a person, then you must be able to answer the question of “You banned this person, why did you not ban that person?”

6. My sense is that the main social media companies cannot give good answers to the question of “You took down X, why did you not take down Y?” A big part of the answer is that the enforcement of terms of service is costly. That is my excuse for not getting rid of every single personal attack in the comment section. But apart from that, I hope that my enforcement is not selective.

7. But enforcement costs do not excuse selectively shutting out ideas or people you dislike. As with laws, selective enforcement undermines the legitimacy of terms of service.

What I believe now, part 3: against naive revelation

A core problem is that people believe too strongly that there is a right way to manage society, and this right way is known.

Jeffrey Friedman uses the jargon “naive realism” to describe the state of mind in which you believe that your way of viewing the world is accurate. Because I trip over the word “realism” in this phrase, I prefer to call it “naive revelation.”

Naive revelation is the belief that the truth has been revealed to you. Closely related is the belief that the truth has been revealed to others (experts) that you can identify.

Naive revelation leads to faith that the solutions to social problems are easily found. This in turn leads to a belief that public officials who do not implement these solutions are blind or evil.

Public officials react to their failures by saying that they could solve problems if you give them more power. (Just let us implement lockdowns using science, and we will solve the problem of the virus.) Naive revelationists fall for this all the time.

The opposite of naive revelation is understanding that we face strong imperfection. Most of the time, the best we can do is adopt trial-and-error groping. But naive revelation is the dominant attitude.

Which is another reason that populism is not a solution. If you believe in populism, then you suffer from a form of naive revelation.

What I believe now, part 2: strong imperfection

What I mean by “strong imperfection” is that human beings and their societal arrangements are very far from perfect. We are nowhere close to utopia, and we cannot see how to get there.

A major reason for this is lack of knowledge. We know today much more than we knew one hundred years ago. It seems reasonable to expect that in another hundred years, today’s level of knowledge will seem low. If we look at all of the past beliefs that today seem wrong-headed, we should be hesitant to commit to what we believe now. On this topic of what we do not know, check out the econtalk podcast with Russ Roberts and Michael Blastland.

The implications of this are:

1. We should be humble about predicting the consequences of public policies. In an economics textbook, a single “market imperfection” is shown in isolation, with the implicit assumption that everything else is perfect. Under those assumptions, the right tax, subsidy, or regulation can reliably produce improvement.

Most economists are familiar with the “theory of the second best,” which points out that trying to fix one problem, when there are other problems or constraints, can make things worse. This is a useful concept, but it only scratches the surface of strong imperfection.

2. We should welcome trial-and-error learning. The economic and social progress we have made is largely due to trial and error, not central planning. Because of strong imperfection, we know that many flaws and problems still exist. It is likely that solutions will come from trial and error going forward, just as in the past.

3. We should try to limit the number of personal flaws that we see as inexcusable. Both as a society and as individuals we should try to extend tolerance and forgiveness. Given our current state, I do not think we can do away with prisons, but I think we should be aiming in the direction of limiting their use. I also think that we should be reducing the number of “firing offenses” in the work place, not adding to them. As individuals, we should aim to reduce the set of excuses for cutting people off as friends.

4. We should avoid the “nirvana fallacy,” which involves comparing the current state to a perfect state. The most realistic change is likely to be from an imperfect current state to another imperfect state.

5. We should resist becoming Manichean. The motives of opponents are usually not as bad as we are inclined to make them out to be.

What I believe now, part 1: grandparenting

It is a conservative’s nature to believe that society has gone off the rails. I believe that we have gone off the rails by having lost sight of the importance of children and grandchildren. I predict that many people today between the ages of 25 and 40 will find themselves becoming lonely and depressed by age 60 as they see the past as having little meaning and the future as having little purpose.

Note that being an aunt or an uncle can have some of the same satisfaction as being a grandparent. But with fewer siblings these days, becoming an aunt or uncle will be rare.

I believe that grandparents are the happiest people. This is based on introspection and observation. Show me a grandparent who does not love their grandchild.

Although I know happy grandparents who are divorced, grandparenting is more satisfying if you managed to stay married. At worst, a divorce may alienate you from your children. At best, it pretty much forces your children to divide up their visits, so that you get half the time with your grandchildren that you would if you had stayed married.

Nobody in their twenties makes decisions based on a desire to end up as grandparents. That is too far to look ahead. Instead, young people respond to immediate cultural influences.

Consider a repressive culture vs. a liberated culture, or R vs. L.

In an R culture, sex outside of heterosexual marriage is frowned upon and difficult to obtain; abortion and birth control are frowned upon and difficult to obtain; and divorces are frowned upon and difficult to obtain. In an L culture, none of these apply.

We are in an L culture, and even if we wanted an R culture there is no squeezing that toothpaste back into the tube. But in an R culture more people are likely to become grandparents.

As an aside, perhaps an L culture is somewhat self-limiting. Imagine that there is a polygenic score that measures likely deviation from straight heterosexual preferences. A score close to 0 means you are very straight. A score close to 10 means you are very non-straight. In an R culture, people with high scores are pressured to conform, so they try to get by in traditional marriages and have children. In an L culture, only people with low scores are likely to have traditional marriages and children. It seems to me that this would lead to a gene pool that tends to reduce the proportion of children who are born with a propensity to deviate from straight heterosexuality.

As another aside, perhaps some other people also want to take themselves out the gene pool.

The official Black Lives Matter organization, which has received vast sums in corporate funding, has listed the abolition of the family among its demands. Left-wing publications like the slickly produced anarchist Commune magazine have explicitly advocated for it. Last year Verso Books, the influential leftist publishing house, released Sophie Lewis’s Full Surrogacy Now: Feminism Against the Family, which made the case for family abolition amid glowing reviews in Vice, the Nation, the Outline, and elsewhere. The Open Society Foundation and Ford Foundation-funded publication Open Democracy recently published an opinion piece by Lewis headlined “The Coronavirus Crisis Shows It’s Time to Abolish the Family,” while the Nation ran with “Want to Dismantle Capitalism? Abolish the Family.”

I believe that grandparents are the most socially forward-looking people. They want the best for their grandchildren. For the most part, I think that this means that grandparents will vote wisely. But when it comes to Social Security and Medicare, I suspect that most grandparents think, “Those entitlements mean that I won’t be a burden on my grandchildren, so it’s good for them,” even though in the aggregate this is not the case.

People who are not raising children or who have never raised children have very little stake in the future of society. Perhaps they should not vote? Raising children means being in the same household with them. So perhaps single mothers should vote, but absent fathers should not?

My intellectual influences, part 5: Net-heads

In the winter of 1993, a group of us at Freddie Mac visited snowbound Albany, New York, to meet with some researchers at General Electric about their automated underwriting project. But while higher-ups were conferring, one of the nerds took me down to the basement to show me Mosaic, the first graphical browser for the World Wide Web. At that point, I became a net-head. Continue reading

Is Final-five voting the answer?

I review the proposal.

Gehl and Porter spell out what they see as the desired state of Congressional politics. This would include these characteristics:

—Effective solutions. These are centrist policies that deal with problems, not perfectly but effectively and with bipartisan support.
—Action. Partisanship should not produce gridlock in the face of difficult issues.
—Broad-based buy-in over time. This means that major legislation would have bipartisan support, rather than being rammed through by the party that happens to hold a majority.
—A balance of short- and long-term needs. Too often, the only bipartisan legislation adds to fiscal deficits and unfunded liabilities.

Whatever the risks are with its adoption, final-five voting should not be compared with some political nirvana. Instead, it ought to be evaluated relative to the current situation, and from that perspective it seems like a promising reform.

Cultural contradictions

Geoff Shullenberger writes,

Taken as a whole, [Christopher] Lasch’s body of writing offers an account of the limitations of the American political panorama of his era. Conservatism, he suggests, tends to provide de facto ideological cover for the economic developments that have eroded the social values it claims to promote. Liberalism, for its part, has overseen the rise of a state bureaucratic apparatus that promises to compensate for the effects of this erosion. However, in the process, it further weakens the autonomy of individuals, families, and communities, and enables the substitution of democracy with technocratic elite rule. While the New Left of the 1960s rebelled against the expansion of corporate and bureaucratic power, the end result of its revolt was not a reassertion of the local and the communal, but the infusion of those structures with a new therapeutic sensibility.

Over the past year, the COVID-19 pandemic has exacerbated many of the economic, social, and cultural tendencies Lasch deplored. While small businesses have collapsed in record numbers and average families have found themselves destitute, the major tech companies and retail conglomerates like Amazon have reaped massive profits and the stock market has soared; in response, the political class has delivered aid packages that blatantly favor the interests of the latter. As Alex Gutentag recently argued, “the pandemic is a convenient scapegoat for the largest upward wealth transfer in modern human history.” Lasch’s work suggests that the roots of this crisis extend far back into the last century, during which both liberals and conservatives, for different reasons, became increasingly indifferent to the degradation of average people’s lives and livelihoods. He offers us no easy alternatives, but his writings reveal the scale of the problems anyone attempting to look beyond the failings of liberalism must confront.

Over the past 250 years we have gone from nation of yeoman farmers to a nation of industrial workers to a nation of white-collar workers in technology, government, and the non-profit sector. With each transformation, the sense of being able to determine our own fate declines, and the sense of dependence on those with concentrations of wealth and power increases.