Links to Consider, 12/19
Sam Hammond on the near-term impact of AI; Rob Henderson on human imitation; Brian Chau and Michael Gibson; James McLeod on the Internet and our minds
education seems especially exposed to disruption by near-term large language models, starting at the youngest ages. Parents are told to read and talk to their kids from infancy, helping prime their language faculty and build some basic vocabulary. So why not have a plush chatbot teddy bear that converses with the child 24-7? It could even teach the infant a second language through repetition and immersion, becoming their very best amigo.
And your young lady’s illustrated Primer won’t be feeding your child Critical Race Theory unless you ask it to! By the way, Hammond is really moving up the charts on the list of substacks that I follow.
Humans haven’t been successful because we are innovators. Rather, we are successful because we don’t think for ourselves, and save time and energy by copying others. Especially those our community deems prestigious. In the ancestral environment, people obtained prestige through wisdom, skill, and experience. Thus, we tend to believe that prestigious people are more competent; prominence is a heuristic for skill.
If course, we are also innovators. Trial-and-error enables us to adapt to new problems and produce new achievements. But for most of us, most of the time, copying is a better strategy than trying to solve a problem by ourselves. Rob Henderson has been high on my charts for a while now, and he keeps turning out the hits.
Henderson warns not to over-imitate the most successful people.
has Michael Gibson on a podcast. I recommend the whole thing, but especially minutes 52-55 on whether talented people are held back by not being discovered or by their own choices to take low-risk paths. Also minutes 1:04 - 1:06 on the “insider-outsider” dynamic. At 1:16 -1:21 Chau’s hypothesis that most people do not make cost-benefit analysis. At 1:37 - 1:43 on valorization of non-profits and the role of envy in hostility to markets. At minute 2:15 - 2:23 Chau argues that machines can be relatively good at things that we think of as “human” skills, like bedside manner. Some riffs about Richard Reeves. writes,If you are an extremely successful author, you don’t have to self-promote your writing anymore. You can wait for others to share it and simply retweet or re-post their endorsements. Some don’t even do that. Some writers are so well known that, despite having millions of followers, they literally don’t promote anything they write on social media. That is some strong countersignaling.
Countersignaling a poor strategy for new writers (or podcasters, or musicians, or others in creative domains). People just starting out should look at those who are a little ahead of where they want to be. They’ll usually find that novice writers who are accruing some success regularly post their stuff online and ask others to share it.
If the predominant form of the printed word is the abstract ideas and arguments, and the predominant form of TV is images and trivia, then I think the predominant form of the internet is mobs, and the idea of an environment driven by metrics.
…Under every social media post and YouTube video, there are metrics — likes, comments, subscribers, views, etc. And in more indirect ways, too, we move in mobs. Memes, aesthetics, fandoms, trending topics and five-star product ratings, are all signals that show us the crowd dynamics online.
Some of it is down to design choices by tech companies, and some of it is maybe just the fundamental nature of a global communication network. When we can connect and communicate online, we will inevitably clump together into groups. And the resulting mobs aren’t passive; they grow, adapt, clash, and fracture internally.
Howard Rheingold, an early Internet enthusiast, wrote a book called Smart Mobs. In an article in 2002, Rheingold wrote,
the most effective use of communication and computer technologies could emerge from new scientific understandings of human cooperation. The most powerful opportunities for human progress are rooted not in electronics but in understandings of social practices. Sociologists, political scientists, evolutionary biologists, even nuclear warfare strategists have contributed the first clues that an interdisciplinary science of cooperation might be emerging.
Mobile communications and pervasive computing have the potential for magnifying cooperation far more powerfully than previous technologies; coupled with new knowledge about the social dynamics of collective action, smart mob technologies could make possible improvements in the way billions of people live.
So far, this is not working out well, as I imagine Rheingold would admit. McLeod, channeling Neil Postman, suggests that the pre-electronic written word produced a healthier society.
"And your young lady’s illustrated Primer won’t be feeding your child Critical Race Theory unless you ask it to!"
That depends on who programmed it!
Henderson's advice to imitate those one rung up from you on the ladder is good for those one rung up, too: it gives them the useful discipline of being a role model for someone they can still easily empathize with, because the memory of being at that stage is still fresh.
This is why Lancastrian teaching is so good and important to do more often. Not only does teaching what you just learned cement the mastery of it in your own mind, but you may be a more effective teacher of the material than someone who mastered it long ago, because you still remember what it was like not to know that material, and the old master may not.