Attending two parallel workshops,
"AI Reasoning in Theoretical Physics" and
"String Theory and the Observable Universe"
at the Aspen Center for Physics.
On Saturday, we went on a hike to American Lake.
There was still some snow up the
Kumiko Kotera, former Burke Fellow at Caltech, is featured in Le Monde discussing her role as Director of the Institut d'astrophysique de Paris.
https://t.co/n2buYITaxe
@karch_andreas@ZoharKo@JosephPConlon@stonybrooku As I pointed out in the other thread, the list is close the classic Zipf law, 1/rank. It is what one expects from preferential-attachment dynamics in the sense of Barabási–Albert: academic careers move through a network in which existing prestige tends to attract new links.
I asked ChatGPT,
"Under what conditions, should I expect a 1/(rank)^{1/2} behavior for current affiliations of invited speakers of a conference?"
Here is what it said:
1. The committee is selecting from a field in which scientific strength is distributed across many institutions, rather than dominated by a few. Then even if the top places still do a bit better, the decline with rank is gentle.
2. The committee is trying, explicitly or implicitly, to balance quality with breadth. Even without formal quotas, a committee often avoids inviting too many people from the same few institutions. That soft “anti-concentration” effect flattens the ranking.
3. Current affiliation is only loosely correlated with long-term prestige. In many fields, people are trained at a few elite hubs but then disperse widely as faculty. If the committee looks only at current institutions, that dispersal naturally makes the current-affiliation curve flatter than the pedigree curve.
It was I who asked ChatGPT to generate this list, prompted by a conversation with Daniel Harlow on Faceboon. It showes the classic Zipf distribution of 1/rank. Note that the list is based on past and current affiliations of invited speakers. For comparison, the list based on current affiliations, attached here, shows a flatter 1/(rank)^{1/2}.
The Caltech group knows it has years of work ahead before any of its dreams have a chance of coming true. But the researchers can’t wait to get started. “Pick a cooler life quest than building the world’s first quantum computer with your friends!” said a jubilant @DolevBluvstein [CEO of @TeamOratomic ], reached by phone shortly before their paper went live, before rushing off to celebrate.
https://t.co/4WHulKAPuK