iiinova — where your digital self keeps evolving

Reposted by Yannick Dupraz

Josh Hadro @hadro.bsky.social · 01/10/2026

Fun fact: I also posted this on LinkedIn, but over there it has escaped containment and I'm now getting a stream of tech industry dudes sharing their one-shot VLM output and saying all it needs is "a good prompt." 🙄

1114

Yannick Dupraz @ydpz.bsky.social · 22/07/2026

I think if you narrow on economic history, it might be 1000 for 1

020

Reposted by Yannick Dupraz

Cédric Chambru @cedricchambru.bsky.social · 07/07/2026

Very happy to share again the CfP for the next edition of the Virtual Workshop on French Economic History. Please submit your paper before August 30, 2026. Submission form: forms.gle/3jYBXmBzF4WL... More information at: sites.google.com/view/vwfhe

167

Reposted by Yannick Dupraz

Christian Vedel @christianvedel.bsky.social · 30/06/2026

Machine Learning is causing massive leaps everywhere. Also in Economic History. There is a conversation we need to have. What ought to be the new scientific standards? Together with two coauthors . Our argument: We should maybe care a bit less about accuracy 📝🧵

41113

Yannick Dupraz @ydpz.bsky.social · 16/06/2026

Joan Miró museum is amazing if you like Miró! (if you don't it's a bit disappointing)

000

Yannick Dupraz @ydpz.bsky.social · 30/04/2026

It's funny how this LinkedIn post proposing to fight bad AI slop was clearly written by chatGPT. The short sentences, the ternary rhythm, the semi-columns before the big reveal: it's all there. Even the overdose of verbless clauses. I want long sentences, broken grammar, confusing thoughts.

020

Yannick Dupraz @ydpz.bsky.social · 02/04/2026

Very happy to have modestly contributed to this work!

020

Yannick Dupraz @ydpz.bsky.social · 04/03/2026

It seems to me, whether we make it official or not, things will slowly start working this way: editors will only desk accept papers that they saw presented at a conference/seminar. We might as well make the rules of the game transparent to all.

000

Yannick Dupraz @ydpz.bsky.social · 04/03/2026

Then of course it introduces a big barrier to entry (like a high submission fee), detrimental to researchers based in low income countries. But you could imagine regional conferences, like the ones the Econometrics Society organizes.

210

Yannick Dupraz @ydpz.bsky.social · 04/03/2026

There are scientific fields (like computational linguistics) where conference proceedings are the main way of disseminating research. It seems like a good filter for slop, as you can verify that the researcher presenting understands the content of the paper presented.

210

Reposted by Yannick Dupraz

Zach Weinersmith @zachweinersmith.bsky.social · 06/01/2026

So, something we got into a bit for the space book is the idea of resources as the source of wealth, because (a) it's widely believed, and (b) it's a claim used to justify all sorts of outlandish space ideas. The short version is that resources are a teeny tiny fraction of wealth.

519357

Reposted by Yannick Dupraz

Thomas Dietterich @tdietterich.bsky.social · 28/12/2025

I'd like to propose the following norm for peer review of papers. If a paper shows clear signs of LLM-generated errors that were not detected by the author, the paper should be immediately rejected. My reasoning: 1/ #ResearchIntegrity

411628

Yannick Dupraz @ydpz.bsky.social · 17/10/2025

I just reviewed my first LLM written paper, with prompts left in the text and invented citations. LLMs are making the (unpaid) job of reviewer much harder. Of course, I could just have the paper reviewed by an LLM, and me and the authors and the other reviewers could go have a drink on the beach!

020

Yannick Dupraz @ydpz.bsky.social · 08/10/2025

New working paper out: A Century of Language Barriers to Migration in India, with Latika Chaudhary and James Fenske. warwick.ac.uk/fac/soc/econ...

011

Reposted by Yannick Dupraz

Dr. Nick Posegay @nposegay.northsky.social · 17/09/2025

I'm sorry, worldwide, irrevocable, non-exclusive, transferable permission to my voice and likeness? For what now? In any manner for any purpose??? This is in academia/.edu's new ToS, which you're prompted to agree to on login. Anyway I'll be jumping ship. You can find my stuff at hcommons.org.

By creating an Account with Academia.edu, you grant us a worldwide, irrevocable, non-exclusive, transferable license, permission, and consent for Academia.edu to use your Member Content and your personal information (including, but not limited to, your name, voice, signature, photograph, likeness, city, institutional affiliations, citations, mentions, publications, and areas of interest) in any manner, including for the purpose of advertising, selling, or soliciting the use or purchase of Academia.edu's Services.

571671851

Yannick Dupraz @ydpz.bsky.social · 13/08/2025

I am looking for a specialist in the Akkadian language. It is for a good cause: a friend has a daughter called Ishtar after the Akkadian goddess (Inanna in Sumer). He wants to confirm the graphic symbols in Akkadian for Ištar. (Yes it is for a tattoo, to avoid tattooing "lentil soup" in Akkadian).

010

Yannick Dupraz @ydpz.bsky.social · 25/06/2025

All this to say that Citizen and Subject is a must read for anyone interested in the economic development and economic history of African countries. press.princeton.edu/books/paperb...

press.princeton.edu

Citizen and Subject

030

Yannick Dupraz @ydpz.bsky.social · 25/06/2025

I saw that Mamdani was in the news and initially though "oh! Ugandan political scientist Mahmood Mamdani is in the news!", but then I realized that the Mamdani in the news was Zohran Mamdani, American politician. But then I checked and one is the son of the other

100

Yannick Dupraz @ydpz.bsky.social · 20/03/2025

The third: it is clearer what the paper is about

020

Yannick Dupraz @ydpz.bsky.social · 05/03/2025

#StandUpForScience Les attaques contre la science aux Etats-Unis ont déjà des répercussions sur la science en Europe, et pourraient préfigurer ce qui nous attends. Il y a des événements organisés un peu partout en France 👉https://standupforscience.fr/tous-les-evenements/

013

Reposted by Yannick Dupraz

Stephanie Kennedy @wordswithsteph.bsky.social · 01/03/2025

As of today, a new term has entered our language. A “rubio” is a spineless lump of nothing. Example: Don’t be such a rubio.

30596735514392

Yannick Dupraz @ydpz.bsky.social · 24/01/2025

Beautiful pictures! and a brilliant weather. My mum's hometown. Do take the Eggs to the Bastille if you have time, astonishing view. If you have even more time, you can hike down back to town.

000

Yannick Dupraz @ydpz.bsky.social · 20/01/2025

Thanks! Writing to you in dm!

000

Yannick Dupraz @ydpz.bsky.social · 20/01/2025

The problem is very common in economic history. I assume it is also common in many other fields. I'd love to see the different approaches people have to tackle the problem. Please RT! Thanks! 3/3

110

Yannick Dupraz @ydpz.bsky.social · 20/01/2025

So for example you are working with data for a country, organized in admin units that are slightly different in each year because some units are split, and other are united. Often you can find an aggregation giving ou a consistent map in all time years. 2/3

110

Yannick Dupraz @ydpz.bsky.social · 20/01/2025

Dear Bluesky community, I am looking for examples of empirical papers working with data organized in changing geographical admin units in two (or more) time periods and needing to aggregate the data up to consistent geographic units, in order to build panel data. 1/3

110

Yannick Dupraz @ydpz.bsky.social · 08/01/2025

Thank you so much for the reference!

010

Yannick Dupraz @ydpz.bsky.social · 08/01/2025

Picture a middle aged man waking up one morning in sweat, realizing he knows practically nothing on the history of China before the 20th century. What books should this man read first?

121

Yannick Dupraz @ydpz.bsky.social · 21/11/2024

In a similar spirit, though not in metrics, there is the Unjournal @unjournal.bsky.social founded by @daaronr.bsky.social

010

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

In the end, French colonialism was cheap, except perhaps in the very last decades, mostly because of independence wars, and colonies received very little public aid.

020

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

The soldiers sent by France to fight in Indochina & Algeria received wages paid by the French government. They used their wage to consume goods, a large share of which were imported. This alone goes a long way into accounting for the trade deficits of French colonies.

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

We show that the deficits were compensated not by French public or private transfers of capital, but by military expenditure spent locally in the colonies, notably during the independence wars.

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

But, Jacques Marseille would reply, what about the trade deficits? how were they compensated? For the paper, we (painfully) tried to reconstruct as many items as we could of the balance of payments of French colonies.

120

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

If we use the OECD definition of aid to compute the development aid given by France to its colonies, we find a figure of 0.21% of GDP on average. This is lower than French development aid today (0.55% of French GDP), and less than a third of the 0.7% UN target.

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

On this graph you can see the striking difference between the pre- and post-WW2 period, and also that most of the cost was for the military.

Graph showing civil subsidy and military expenditure on the colonies as a percentage of French GDP

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

Results: France spent very little on its colonial empire: 1.3% of its GDP on average, of which 82% was for the military. Between 1833 & 1939, the cost was even lower, 0.5% of GDP, but the cost surged to 3.2% of GDP between 1946 and 1962, mostly because of decolonization wars.

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

We delved into the archives to directly calculate how much France spent on its colonial empire between 1830 (beginning of the conquest of Algeria) and 1962 (independence of Algeria).

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

The balance of payment must always be in equilibrium (this is just an accounting identity), so if the trade balance is in deficit, something else must be in surplus. For Marseille, this had to mean the colonies were receiving public aid and capital.

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

Marseille's argument relied chiefly on the analysis of the trade balance of French colonies, which was in structural deficit. Marseille argued that this deficit was compensated by important transfers of public and private capital.

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

According to Jacques Marseille in his 1984 book, colonization was a huge burden on France's public finances that grew overtime and became unsustainable on the eve of decolonization.

110

Yannick Dupraz @ydpz.bsky.social · 13/06/2024

New Working Paper: "Colonialism on the Cheap: The French Empire 1830-1962”, with Denis Cogneau, Elise Huillery & Sandrine Mesplé-Somps. French colonies did not cost French taxpayers much. Most of the cost was for the military, especially for the independence wars. shs.hal.science/halshs-04598...

shs.hal.science

Colonialism on the Cheap: The French Empire 1830-1962

How much did France pay for its colonial empire? Did colonies benefit from large transfers from French taxpayers and private investors, or were they on the contrary drained of their capital? So far, J...

121

Yannick Dupraz @ydpz.bsky.social · 20/05/2024

Interesting point. I might be wrong, but I feel that in order to ask chatgpt the right questions and in order to be able to judge the quality of the answer provided, you'd need to be quite familiar with the process of data cleaning and data manipulation.

100

Yannick Dupraz @ydpz.bsky.social · 20/02/2024

Sorry, just saw you have an earlier thread specifically on this!

010

Yannick Dupraz @ydpz.bsky.social · 20/02/2024

Thank you! Great to have you lay down your thoughts on this!

010

Yannick Dupraz @ydpz.bsky.social · 20/02/2024

Hello and thanks for the blueskying! What do you make of this meta analysis of returns to education by Clark and Nielsen? They claim there is pub bias on sign www.ehes.org/wp/EHES_249....

100

Yannick Dupraz @ydpz.bsky.social · 05/02/2024

A summary of our recently published work with Yasmine Bekkouche on colonial origins and quality of education in Cameroon www.aehnetwork.org/blog/colonia...

aehnetwork.org

Colonial Origins and Quality of Education: Evidence from Cameroon – African Economic History Netwo...

011

Yannick Dupraz @ydpz.bsky.social · 02/02/2024

True! Also, I think in other fields, like biology, the creators of the dataset tend to become, if they want to, co-authors of all subsequent papers using the data. They become middle authors (as opposed to first or last), which amounts to little, but still it increases the returns to data creation.

010