Assistant Professor, Economics (I.O., Digital, and Labor) at @Telecomparis (CREST)
Télécom Paris
Joined September 2018
- Tweets1.1K
- Following4K
- Followers1.2K
- Likes10.4K
Pinned Tweet
⚠️ New paper alert ⚠️
We study the Digital Markets Act Impact on Google Maps with @millo_rossi (cesifo.org/en/publications/2… )
A thread 1/n 🧵
The full draft cowritten with Michelangelo Rossi and Adèle Diehl is available here : ifo.de/DocDL/cesifo1_wp13098…
All in all, this paper shows that distribution matters for AI competition.
This may be relevant when thinking about self-preferencing behavior in the context of the DMA.
This diversion did not come from users searching less for ChatGPT. Rather, we find that distributing an AI chatbot on a platform users are already active and visiting intercepts traffic, directly.
We study the August 2025 rollout of AI Mode to 30 countries, using countries who did not yet receive AI Mode as a control group.
We find that ChatGPT lost around 6% of traffic, equivalent in volumes to 70% of the traffic going to Gemini's standalone traffic.
Note that Google is not the only to do so. Grok within X is also a case of such integration. Meta AI in WhatsApp... Copilot in Windows... the list goes on
⚠️New working paper⚠️
We study Google's integration of AI Mode, a conversational chatbot, within Google Search.
Although Gemini already existed as a standalone AI, you now have a chatbot directly on a website you were already visiting.
How did it affect AI competition?
Louis Pape retweeted
Happy to share the latest version of my paper “Scalable Nonparametric Demand Estimation in Differentiated Product Markets” tinyurl.com/bdd3u999
Louis Pape retweeted
👏
Avec sa fille Christine et la Fondation qui porte son nom, j’ai dévoilé au 83 rue Didot - l’immeuble où il est né- une plaque en hommage à celui qui fut notre premier Prix Nobel d’économie, Maurice Allais.
Pupille de la Nation, orphelin issu d'un milieu très modeste, Maurice Allais a montré toute sa vie que le déterminisme social n’est pas une fatalité.
Jusqu’à la plus haute reconnaissance internationale, tout son talent a pu éclore et ses efforts et son mérite porter leurs fruits grâce à l’École de la République qui lui a tendu la main, de l’école primaire de la rue d’Alesia à @Polytechnique.
Toute sa vie, il a proclamé cet impératif républicain d’offrir à chacun le même point de départ.
#MauriceAllais #Paris14 #FondationMinesParis @Mines_Paris
Louis Pape retweeted
Ce matin, Google a allumé AI Overviews en France. Deux ans qu’on regardait les autres pays perdre 20 à 40 % de leur trafic.
Je ne suis ni pro ni anti-IA. Je suis pro-contenu. Et une machine qui résume tout sans rien créer finira par résumer du vide.
Je suis en congé paternité depuis une semaine, mais je reste en contact avec l’actualité de mon métier. Entre deux changements de couches 😅
La France était le dernier grand marché épargné. On a donc eu deux ans pour observer ce qui se passe ailleurs : des baisses de trafic de 20 à 40 % chez les éditeurs, et jusqu’à -58 % de clics sur la première position quand un résumé IA s’affiche. Ce ne sont pas des projections. Ce sont des données réelles.
Le débat n’est pas « pro-IA » contre « anti-IA ». Ce clivage est une impasse. L’intelligence artificielle a plusieurs formes, j’en utilise tous les jours, dans nos rédactions, dans nos outils. Certaines augmentent le travail des journalistes. D’autres se contentent d’aspirer ce travail pour le restituer sans que personne ne visite jamais la source.
Le vrai sujet, c’est la protection du contenu. De la création. Un résumé IA n’existe que parce que quelqu’un, quelque part, a enquêté, testé, vérifié, écrit. Si ce quelqu’un ne peut plus vivre de son travail, il n’y aura plus rien à résumer. C’est aussi simple que ça.
Google promet un opt-out, de la transparence, des droits voisins. Très bien. Mais un opt-out qui vous fait disparaître, ce n’est pas un choix. Pour moi, c’est du chantage tout simplement.
Il faudra se battre pour la création, non contre la technologie, mais pour ceux qui conçoivent ce qu’elle consomme.
I've read that AI makes interns & RAs redundant. But I've found that interns & RAs have become a lot more productive too! (or maybe I was just lucky with my recent hires)
Louis Pape retweeted
Something I've observed in academia but I suspect is true in industry as well: as you become more senior, you'll find that you can be much more productive if you spend all your time being a manager and letting junior people do all the real work. But after a few years of this, you're likely to lose your skill at doing the work yourself, the opportunity to practice that skill, and the gratification of doing so that drew you to your chosen profession in the first place. And you'll probably also get worse at being a manager as you lose touch with the actual work.
Avoiding managerdom is hard. You have to give up substantial short-term productivity gains. If you're in academia in a tenure track, you risk not getting tenure by being less productive. And personally I find that switching between thinker mode (deep work with no distractions for a whole day) and manager mode (a thousand quick but urgent tasks) to be highly unpleasant. When I'm in thinker mode, it kills me that people have to wait a day (often a lot more) for a 1-minute response from me that could unblock them on what they're stuck on and save them hours of work. When I'm in manager mode, it kills me that I have all these creative ideas sloshing around in my brain that I'm not able to execute on.
So it's extremely tempting to perpetually be in just one mode or the other. When you have a team, not managing them is not an option, but being a pure manager is an option. That makes a constant struggle to carve out time for your own work and not give in to the temptation.
Note: I think this is only tangentially related to Founder Mode vs Manager Mode. It's much more related to another Paul Graham essay, Maker's Schedule vs Manager's Schedule. And Deep Work by Cal Newport is very relevant.
General take-away which, I think, speaks more generally about the problems related to income inequality : when something is designed with only a subset of the user base in mind, the product is not likely to be well designed for the remaining users. This is a problem. N/N
