Bernard Pivot left a lasting impression on me for a somewhat forgotten project
A beloved French TV personality, Bernard Pivot passed away two weeks ago at 89. His literary shows, including his flagships "Apostrophes" and "Bouillon de Culture", were cultural gems, which inspired James Lipton's "Inside the Actors Studio". But the real reason I was so fond of him was because of a prank he orchestrated on a national scale in 1996. At the time, audience measurement consistently pushed his shows into the late night: cultural shows do not seem to rank high in any country in the world. But being forced into an undesirable slot creates a self-fulfilling prophecy, as few people are willing to stay up until 11pm to watch any show. So he decided to strike back with his own self-fulfilling prophecy.
His ploy comprised writing, producing and airing a TV movie “Coeur de Cible” (Target Audience). The movie tells the story of a serial killer whose victims are all... panel members for an audience measurement company. When the movie aired in 1996, the real-life counterparts of his fictional set of victims were disproportionately riveted to the film, thus artificially skewing ratings and throwing the audience measurement industry into temporary disarray! Audimat, the main French audience measurement and ratings firm was appalled (and vocal about it) as it was unable that day to provide statistically reliable audience estimates. Bernard Pivot had its revenge.
What are the most influential papers on agent-based modeling (ABM)?
Not only did a certain 2002 paper (link) make the list, it turns out to be the most cited article, with more than 6,000 citations as of May 11th, 2024. There is no exciting breakthrough in that paper, no revolutionary discovery. It did, however, provide an easy framework for thinking about the benefits (and costs) of ABM, when ABM makes sense, when it doesn't. In the couple of years after its publication, I heard from a lot of people that it been inspiring to them.
A lot of things have happened since 2002. Estimation and calibration of nonlinear interacting systems. Computing power. And with hashtag#AI and hashtag#genAI, a whole new set of techniques have become available to accelerate progress and, I hope, adoption.
Now might be a good time for an updated version of that paper. Doyne Farmer jean-philippe bouchaud, let's do it!