Links to Consider, 1/5
Noah Smith on AI; Infovores links from 2022; Jonathan Haidt on kids today; Freddie deBoer on Haidt
A big knock on AI is that because it doesn’t really let you understand the things you’re predicting, it’s unscientific. And in a formal sense, I think this is true. But instead of spending our effort on a neverending (and probably fruitless) quest to make AI fully interpretable, I think we should recognize that science is only one possible tool for predicting and controlling the world. Compared to science, black-box prediction has both strengths and weaknesses.
One weakness — the downside of being “unscientific” — is that without simple laws, it’s harder to anticipate when the power of AI will fail us.
As usual, I recommend the whole essay. He suggests that machine learning could turn out to be as important a development as the scientific method. As you know, I am less impressed with machine learning. Because the GPT software gives human-like responses, we attribute intelligence to it that is beyond what it really possesses. We intuitively believe that there is a mind at work, even if the information it is generating has accuracy that is at the same level, or even lower than, that provided by Google search.
Humans think in terms of metaphors. See Steven Pinker’s book The Stuff of Thought. When a computer comes up with an original metaphor that offers insight, I will be willing to call that artificial intelligence. Not what I have seen so far.
is at it again, offering a themed set of book recommendations. Think of these as syllabi for seminars. links of the year, 2022.In the WSJ, Tunku Varadarajan talks with Jonathan Haidt.
he has in mind the younger cohort, Generation Z, usually defined as those born between 1997 and 2012. “When you look at Americans born after 1995,” Mr. Haidt says, “what you find is that they have extraordinarily high rates of anxiety, depression, self-harm, suicide and fragility.” There has “never been a generation this depressed, anxious and fragile.”
…University students who matriculated starting in 2014 or so have arrived on campus in defend mode: “Here they are in the safest, most welcoming, most inclusive, most antiracist places on the planet, but many of them were acting like they were entering some sort of dystopian, threatening, immoral world.” Once they enter the workplace, they’re less innovative, less inclined to take risks, and that may “undermine American capitalism,” Mr. Haidt says.
Have a nice day.
writes,Because Haidt talked about a culture of victimhood, he was immediately coded as right-wing, which is to say on the wrong side of the culture war. (That’s the culture war that sees Elon Musk as a bigger target than Mitch McConnell and Joe Rogan as more dangerous than the CEO of Goldman Sachs.) And so Haidt’s concerns about depression and anxiety among Gen Z, which would seem to be an issue of great public interest, were immediately drowned out by the sound of a thousand bitter people dunking and loling. Of course, Haidt isn’t helping himself any. The term “culture of victimhood” reminds many people of the “snowflake” insult
Freddie notes that social media tend to discourage taking nuanced “on the one hand. . .on the other” points of view. I would say that this is true of Twitter, but it is not true of Substack. Long live Substack.
Re: "Here they are in the safest, most welcoming, most inclusive, most antiracist places on the planet, but many of them were acting like they were entering some sort of dystopian, threatening, immoral world.”--Jonathan Haidt
There is a major blind spot in safety on campus: Sex.
There really are dystopian, threatening, and immoral dimensions to the battle of the sexes on campus.
The problem rests on 3 empirical premisses:
1. Young women are more likely than young men to seek a monogamous relationship. (At most residential campuses, this kind of mismatch is exacerbated by imbalance in the sex ratio on campus, where women substantially outnumber men.)
2. What students call "going out" -- the long-weekend, nocturnal party-and-dating scene -- is awash in alcohol. Why need a drug? Why alcohol, more than other readily available mind drugs? Because alcohol *disinhibits.* Many students drink in order to "remove the mask." (Perhaps status anxiety would otherwise inhibit their sociability or openness.)
3. Sex then usually takes place in a private setting (a dorm room or frat house bedroom), behind closed doors. If something goes wrong in the encounter -- whether coercion during the encounter, or murkiness about consent because of drunkenness, or deep "morning-after" regret when sober -- it is hard to establish the facts. "He said, she said."
There are strong norms among students against sexual predation, but norm-enforcement is often thwarted by intractable information problems. Rumors circulate. Students often don't know what or whom to believe. Moreover, students mistrust the motives/competence of "the machine" (official campus investigation of complaints). They are caught (or ensnare themselves?) in no-man's land, where neither norms nor authorities can reliably provide safety or remedy.
An irony is that this real blind spot in safety occurs also because students believe, per college ideology, that safety is an entitlement, regardless of risk-taking. This belief lowers the guard and provides cover for heavy drink.
It is remarkable that elites send their offspring into the moral mess of the battle of the sexes on campus. Perhaps elites feel trapped when they endorse and fund, for their daughters and sons, the college mix of formal credentials, peer matching in romance, delay of adulthood, and rather risky rites of passage.
"Because Haidt talked about a culture of victimhood, he was immediately coded as right-wing." Imagine what it feels like to be the guy he was talking about when he brought it up (he was talking about an article by myself and Bradley Campbell). FYI, I am employed in a sociology department.
To be fair, I am usually in the lower right quadrant of the compass ( I subscribe to Kling, after all), and Haidt probably isn't. Haidt also seems to prefer the term "safety culture," which I think misses the idea that it's only safety for designated victim groups (members of designated privileged groups might actually deserve some horrible punishment). Actually, Campbell and I reject "snowflake" for similar reasons, aside from its almost purely moralistic connotation.
To be fair to us, Campbell and I have went on untill we're blue in the face about why we think the label makes sense for comparative purposes (it is juxtaposed to honor and dignity cultures) and isn't meant as a slur (even if we find many aspects of the culture objectionable). I am writing some more about the topic on Substack (e.g., https://jasonmanning.substack.com/p/moral-cultures-2-victimhood)