Yannick Oswald

layout opportunities to engage with and reimagine political, economic and social systems by means of simulating them.

I hope it is interesting and even useful to some.

I am not the first to address the simulation of Utopias, but I am the first to scope the literature, connect the dots across fields as diverse as the sociology of Utopia, economics, complexity science, artificial intelligence and so forth and

Back from holiday with a bang 💫

Am pleased my article Artificial Utopia is now published in the journal Futures. lnkd.in/e_fnEbg3

This article conceptualises thinking about Utopian and radically socially innovative societies with the help of computational simulations.

I believe that if you are interested in modelling social change, also in relation to climate change, postgrowth economics etc. this winter school could be quite useful to you.

schumpeter-centre.uni-graz.at/de/veranstal...

Why the German chancellor is wrong - the focus on increasing working time is misguided - and what to do instead Simply working more will not address Germany's deep-rooted problems

Why the German chancellor is wrong - the focus on increasing working time is misguided - and what to do instead
open.substack.com/pub/theworld...

Why the German chancellor is wrong - the focus on increasing working time is misguided - and what to do instead Simply working more will not address Germany's deep rooted problems

Here is a blog on the current debate around increasing work time in Germany.

Instead, we must think out of the box and experiment with drastic economic and social innovation.

theworldinmodels.substack.com/p/why-the-ge...

The AI Review Lottery: Widespread AI-Assisted Peer Reviews Boost Paper Scores and Acceptance Rates Journals and conferences worry that peer reviews assisted by artificial intelligence (AI), in particular, large language models (LLMs), may negatively influence the validity and fairness of the peer-r...

Interestingly enough there is even a paper on this that finds AI biases the review process towards more positive outcomes. arxiv.org/abs/2405.02150 Perhaps. But I think the final outcome is yet to see and regardless of the sign of the outcome, the quality possibly decreases.

I think using AI to refine reviews as any other text is fine, as long as it remains a high quality honest review, but in my case it went so far that the "reviews" clearly just rephrased limitations from my own text and any AI detector would immediately detect it. I felt cheated.

Michael Black on X: "The use of AI in reviewing is a growing problem. Several of my ICCV papers have AI reviews -- one reviewer was so lazy that they left in the prompts! A common refrain that I hear is that people have difficulty writing in English and need to use AI to clean up their review." / X The use of AI in reviewing is a growing problem. Several of my ICCV papers have AI reviews -- one reviewer was so lazy that they left in the prompts! A common refrain that I hear is that people have difficulty writing in English and need to use AI to clean up their review.

Sorry to link to X, but I am so glad another (senior) person is speaking out about this, it is a different field than mine but I had the same problem with a paper that got rejected and even an ERC grant application.

x.com/Michael_J_Bl...

Intergenerational mobility around the world: A new database Using individual data from over 400 household surveys, this paper compiles a global database of intergenerational mobility in education for 153 countr…

Very interesting new database on intergenerational mobility www.sciencedirect.com/science/arti...

👉AI for good? AI for agent-based models? What kind of forms could this take.

Please note the deadline to our SSC special track this Friday! We are looking forward to your research.

👉Also Please spread if you can!!😀

Discussion welcome!

And lastly I speculate about the future and general ability of computational research to aid in such democratisation and scientific efforts. What are its limits? Ethical propblems?

Further I generally systematize and map simulation approaches to their characteristic domains of validity and specific application areas in studying radical democratisation proposal

More concretely I conceptualise one specific possible model architecture to study an alternative political economy in which citizen-assembly-like deliberation plays a central role in societal decision-making

Specifically, I map exisiting simulation methods to existing challenges in niche democratisation efforts such as citizen assemblies and democratic firms to generate a better understanding of how computational simulation may aid social, political and economic innovation.

Hope you enjoy!✍️

🌳It is a research perspective piece intending to conceive the potential for applying computational simulation methods and artificial intelligence to investigate new and alternative democratic paradigms of political economy.

Artificial Utopia: Simulation and Intelligent Agents for a Democratised Future Prevailing top-down systems in politics and economics struggle to keep pace with the pressing challenges of the 21st century, such as climate change, social inequality and conflict. Bottom-up democrat...

👉🙌 😀I am happy to share my new article (preprint) Artificial Utopia: Simulation and Intelligent Agents for a Democratised Future

arxiv.org/abs/2503.07364

Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researchers Recent advancements in large language models (LLMs) have sparked optimism about their potential to accelerate scientific discovery, with a growing number of works proposing research agents that autono...

But evidence shows different already.

Here is a study showing that LLM generated ideas are judged as more novel than researcher generated ideas in NLP.

arxiv.org/abs/2409.04109

But evidence shows different already.

Here is a study showing that LLM generated ideas are judged as more novel than researcher generated ideas in NLP.

arxiv.org/abs/2409.04109

Crazy that in this whole tumult around Zelensky and Trump, the Americans, Trump and his friends, also just swept under the rug that Russia already allied with North Korea and North Korea literally has troops fighting alongside Russia against Ukrainians.

But sure *Zelensky* gambles with WW3.

One scenario the EU must be prepared for now is the U.S. lifting sanctions on Russia.

Key principles for the EU:
- make sure to stay united
- not return to dangerous dependencies
- not deliberately support the Russian war economy

Not changing anything might not be the right answer

I am reading and learning about models of opinion and cultural change and diffusion.

Particularly with relevance to agent-based models or network models.

The main classes I have identified so far are as follows. Anything I am missing?

I am reading and learning about models of opinion and cultural change and diffusion.

Particularly with relevance to agent-based models or network models.

The main classes I have identified so far are as follows. Anything I am missing?

Has the Dawn of Everything by Graeber and Wengrow be misintepreted by a lot of people?

They actually argue themselves that hierarchical and egalitarian societies are around since complex social life emerged, not that all hunterer gatheres were egalitarian. research-portal.uu.nl/ws/portalfil...

21 % für die AfD sind 21 % zu viel.

@wbarfuss.bsky.social @wbarfuss.datasci.social.ap.brid.gy

Please share if possible with your networks. Thanks a lot!

Exciting News 📣🫵

Wolfram Barfuss and I are organising a special track at the
Social Simulation Conference 2025 in Delft.

On "Artificially Intelligent Agents in Social Agent-based Models"

The call is out now. ssc2025.tbm.tudelft.nl/important-da... Looking forward to exciting contributions.✍️

In the bigger picture, for me, this is about complex system prediction, as evidenced by our discussion on its contribution to areas such as urban scaling theory.