Mehmet Mars Seven ♞
@mehmetmars7.bsky.social
Lecturer @kcl-spe.bsky.social @kingscollegelondon.bsky.social
Game Theory, Econ & CS, Pol-Econ, Sport
Chess ♟️
Game Theory Corner at Norway Chess
Studied in Istanbul -> Paris -> Bielefeld -> Maastricht
https://linktr.ee/drmehmetismail
Views are my own
Game Theory, Econ & CS, Pol-Econ, Sport
Chess ♟️
Game Theory Corner at Norway Chess
Studied in Istanbul -> Paris -> Bielefeld -> Maastricht
https://linktr.ee/drmehmetismail
Views are my own
Pinned
I'm sure many names are missing, but even putting this list together took some time. Please feel free to suggest more names from any discipline, economics, computer science, political science, biology, mathematics, and so on, either via DM or by simply tagging them below
go.bsky.app/RjAhP7Z
go.bsky.app/RjAhP7Z
Simulator's Dilemma: If we fear our simulations could outsmart us, imagine what our simulators (if any) might be thinking right now
November 10, 2025 at 11:58 AM
Simulator's Dilemma: If we fear our simulations could outsmart us, imagine what our simulators (if any) might be thinking right now
Reposted by Mehmet Mars Seven ♞
Econ, artificial intelligence and machine learning …
#EconConf #econsky
marketdesigner.blogspot.com/2025/11/econ...
#EconConf #econsky
marketdesigner.blogspot.com/2025/11/econ...
Economics and CS (AI+ML) in Ithaca in June: call for papers
marketdesigner.blogspot.com
November 9, 2025 at 2:08 PM
Econ, artificial intelligence and machine learning …
#EconConf #econsky
marketdesigner.blogspot.com/2025/11/econ...
#EconConf #econsky
marketdesigner.blogspot.com/2025/11/econ...
"Building a pipeline of .. AI tools.. for the Kakeya problem, AlphaEvolve discovered an interesting general construction. When we fed this .. to Deep Think, it successfully derived a proof.. This proof was then fully formalized in the Lean proof assistant using AlphaProof" arxiv.org/pdf/2511.02864
arxiv.org
November 6, 2025 at 12:13 PM
"Building a pipeline of .. AI tools.. for the Kakeya problem, AlphaEvolve discovered an interesting general construction. When we fed this .. to Deep Think, it successfully derived a proof.. This proof was then fully formalized in the Lean proof assistant using AlphaProof" arxiv.org/pdf/2511.02864
Reposted by Mehmet Mars Seven ♞
lol Cowen is at it again
November 5, 2025 at 8:02 PM
lol Cowen is at it again
Quick take: The NYC election showed how social media amplifies extremes.
Evidence-based debate on real issues eg affordability was missing.
We need politicians, economists and political scientists to debate and work together on solutions and trial them. “Easy fixes” are rarely as easy as they seem
Evidence-based debate on real issues eg affordability was missing.
We need politicians, economists and political scientists to debate and work together on solutions and trial them. “Easy fixes” are rarely as easy as they seem
November 5, 2025 at 5:30 PM
Quick take: The NYC election showed how social media amplifies extremes.
Evidence-based debate on real issues eg affordability was missing.
We need politicians, economists and political scientists to debate and work together on solutions and trial them. “Easy fixes” are rarely as easy as they seem
Evidence-based debate on real issues eg affordability was missing.
We need politicians, economists and political scientists to debate and work together on solutions and trial them. “Easy fixes” are rarely as easy as they seem
(1) Ben Golub seems to have found a promising use case.
I’ve already bought 1 & thought it was worth it.
For anyone curious, try via this referral: If you buy 1, you get 1 free for you & me. I'm mostly interested in comparing it further vs GPT Pro and Deep Think
www.refine.ink?ref=bWVobWV0...
I’ve already bought 1 & thought it was worth it.
For anyone curious, try via this referral: If you buy 1, you get 1 free for you & me. I'm mostly interested in comparing it further vs GPT Pro and Deep Think
www.refine.ink?ref=bWVobWV0...
refine
AI feedback for research.
www.refine.ink
November 4, 2025 at 1:54 PM
(1) Ben Golub seems to have found a promising use case.
I’ve already bought 1 & thought it was worth it.
For anyone curious, try via this referral: If you buy 1, you get 1 free for you & me. I'm mostly interested in comparing it further vs GPT Pro and Deep Think
www.refine.ink?ref=bWVobWV0...
I’ve already bought 1 & thought it was worth it.
For anyone curious, try via this referral: If you buy 1, you get 1 free for you & me. I'm mostly interested in comparing it further vs GPT Pro and Deep Think
www.refine.ink?ref=bWVobWV0...
ChatGPT Pro may be making the news in math, but in #gametheory it still makes basic mistakes -- the same kinds humans often make. Here, it tried to prove a unique equilibrium in an infinite-horizon game just because the stage game had a unique equilibrium 😅
#econsky
#econsky
November 2, 2025 at 1:16 PM
ChatGPT Pro may be making the news in math, but in #gametheory it still makes basic mistakes -- the same kinds humans often make. Here, it tried to prove a unique equilibrium in an infinite-horizon game just because the stage game had a unique equilibrium 😅
#econsky
#econsky
Reposted by Mehmet Mars Seven ♞
I crossed an interesting threshold yesterday, which I think many other mathematicians have been crossing recently as well. In the middle of trying to prove a result, I identified a statement that looked true and that would, if true, be useful to me. 1/3
October 31, 2025 at 7:25 PM
I crossed an interesting threshold yesterday, which I think many other mathematicians have been crossing recently as well. In the middle of trying to prove a result, I identified a statement that looked true and that would, if true, be useful to me. 1/3
Live: King's Experts Series: The Future of Higher Education
www.youtube.com/watch?v=kBYP...
@kingscollegelondon.bsky.social
www.youtube.com/watch?v=kBYP...
@kingscollegelondon.bsky.social
King's Experts Series: The Future of Higher Education
YouTube video by KCL Alumni
www.youtube.com
October 30, 2025 at 6:31 PM
Live: King's Experts Series: The Future of Higher Education
www.youtube.com/watch?v=kBYP...
@kingscollegelondon.bsky.social
www.youtube.com/watch?v=kBYP...
@kingscollegelondon.bsky.social
A simple suggestion to improve email user experience: Each time I click “Reply All” to a mailing list, Outlook and Gmail etc. should display a visible highlight in the message-typing area. The highlight can be removed with a simple acknowledgement from the user
October 30, 2025 at 11:55 AM
A simple suggestion to improve email user experience: Each time I click “Reply All” to a mailing list, Outlook and Gmail etc. should display a visible highlight in the message-typing area. The highlight can be removed with a simple acknowledgement from the user
Reposted by Mehmet Mars Seven ♞
As usual, some surprises in the exit poll for the Dutch elections. D66 largest in the exit poll, just before PVV. Given MOE still uncertain which party has won. VVD more or less stable, and given earlier poll losses, they celebrate that as a win. GL-PvdA lose 5 seats. CDA gains.
October 29, 2025 at 10:21 PM
As usual, some surprises in the exit poll for the Dutch elections. D66 largest in the exit poll, just before PVV. Given MOE still uncertain which party has won. VVD more or less stable, and given earlier poll losses, they celebrate that as a win. GL-PvdA lose 5 seats. CDA gains.
Reposted by Mehmet Mars Seven ♞
ChatGPT will never be forgiven for ruining the em-dash.
October 29, 2025 at 7:01 PM
ChatGPT will never be forgiven for ruining the em-dash.
Reposted by Mehmet Mars Seven ♞
Girls who start with the same age and rating as boys will continue to play chess at a similar rate and at a similar level as those boys, especially in areas where participation is more gender-balanced, according to a new study.
www.chess.com/news/view/li...
www.chess.com/news/view/li...
Study Reveals New Insights Into Gender Gap In Chess: Equal Start, Equal Progress
Girls who start with the same age and rating as boys will continue to play chess at a similar rate and at a similar level as those boys, a new study revealed.
www.chess.com
October 28, 2025 at 6:44 PM
Girls who start with the same age and rating as boys will continue to play chess at a similar rate and at a similar level as those boys, especially in areas where participation is more gender-balanced, according to a new study.
www.chess.com/news/view/li...
www.chess.com/news/view/li...
Via Gil Kalai:
mathoverflow.net/questions/50...
mathoverflow.net/questions/50...
Examples for the use of AI and especially LLMs in notable mathematical developments
The purpose of this question is to collect examples where large language models (LLMs) like ChatGPT have led to notable mathematical developments.
The emphasis in this question is on LLMs, but ans...
mathoverflow.net
October 28, 2025 at 3:07 PM
Via Gil Kalai:
mathoverflow.net/questions/50...
mathoverflow.net/questions/50...
Reposted by Mehmet Mars Seven ♞
PROCJAM - the Procedural Generation Jam - is back for another year! Join us on itch.io this November for another week of making something that makes something - a bskybot, a videogame, a digital artwork, a dice table or something else!
🗓️ Nov 21st - Dec 1st
🔮 itch.io/jam/procjam
🗓️ Nov 21st - Dec 1st
🔮 itch.io/jam/procjam
PROCJAM - The Procedural Generation Jam
A game jam from 2025-11-21 to 2025-12-01 hosted by PROCJAM. Let's Make Something! PROCJAM - the Procedural Generation Jam - is a creative coding jam about making things that make things. This normally...
itch.io
October 27, 2025 at 3:35 PM
PROCJAM - the Procedural Generation Jam - is back for another year! Join us on itch.io this November for another week of making something that makes something - a bskybot, a videogame, a digital artwork, a dice table or something else!
🗓️ Nov 21st - Dec 1st
🔮 itch.io/jam/procjam
🗓️ Nov 21st - Dec 1st
🔮 itch.io/jam/procjam
LLMs + proof assistants such as Lean, have the potential to revolutionize maths once they are efficiently integrated, creating a mathematics engine in the spirit of a chess engine. A golden age of economic theory, and of other mathematical disciplines, may well follow
mathstodon.xyz/@tao/1154202...
mathstodon.xyz/@tao/1154202...
Terence Tao (@tao@mathstodon.xyz)
Yet another recent example of AI assistance in mathematics: my colleague Ernest Ryu here at UCLA was able to solve an open problem in optimization theory (roughly speaking, an asymptotic convergence r...
mathstodon.xyz
October 26, 2025 at 3:14 PM
LLMs + proof assistants such as Lean, have the potential to revolutionize maths once they are efficiently integrated, creating a mathematics engine in the spirit of a chess engine. A golden age of economic theory, and of other mathematical disciplines, may well follow
mathstodon.xyz/@tao/1154202...
mathstodon.xyz/@tao/1154202...
I'm here for posts like 👇
I may not agree with everything (sometimes I totally disagree, like this piece) but I very much value finding links to thought-provoking posts and discussions
thanks @philipgoff.bsky.social
I may not agree with everything (sometimes I totally disagree, like this piece) but I very much value finding links to thought-provoking posts and discussions
thanks @philipgoff.bsky.social
"A Mathematic Proof that God Beats the Multiverse." Who needs philosophy when you get settle it with a definitive proof?
open.substack.com/pub/philipgo...
open.substack.com/pub/philipgo...
A Mathematical Proof that God Beats the Multiverse
Image by https://unsplash.com/@tyvdh
open.substack.com
October 24, 2025 at 4:00 PM
I'm here for posts like 👇
I may not agree with everything (sometimes I totally disagree, like this piece) but I very much value finding links to thought-provoking posts and discussions
thanks @philipgoff.bsky.social
I may not agree with everything (sometimes I totally disagree, like this piece) but I very much value finding links to thought-provoking posts and discussions
thanks @philipgoff.bsky.social
Reposted by Mehmet Mars Seven ♞
And now might be a good time to mention, I’m on the faculty job market this year! I do work in human-AI collusion, collaboration and competition, with an eye towards building foundations for trustworthy AI. Check out more info on my website here! Nataliecollina.com
October 22, 2025 at 3:56 PM
And now might be a good time to mention, I’m on the faculty job market this year! I do work in human-AI collusion, collaboration and competition, with an eye towards building foundations for trustworthy AI. Check out more info on my website here! Nataliecollina.com
Reposted by Mehmet Mars Seven ♞
“How is this economics?”
October 22, 2025 at 11:28 AM
“How is this economics?”
I find "A Definition of AGI" very informative and a contribution in the right direction.
One major missing component is Game Theory, and especially distinction between zero-sum tasks (pure competition) vs. non-zero-sum games (competition & cooperation) www.agidefinition.ai/paper.pdf
One major missing component is Game Theory, and especially distinction between zero-sum tasks (pure competition) vs. non-zero-sum games (competition & cooperation) www.agidefinition.ai/paper.pdf
October 21, 2025 at 12:04 PM
I find "A Definition of AGI" very informative and a contribution in the right direction.
One major missing component is Game Theory, and especially distinction between zero-sum tasks (pure competition) vs. non-zero-sum games (competition & cooperation) www.agidefinition.ai/paper.pdf
One major missing component is Game Theory, and especially distinction between zero-sum tasks (pure competition) vs. non-zero-sum games (competition & cooperation) www.agidefinition.ai/paper.pdf
Math is super helpful but a limited tool. What matters is if research is insightful, not how mathy it is.
Think of chess: try becoming a GM with math models. Its concepts are too abstract, like economy and people, for humans to grasp that way, even if engines show it’s possible
Think of chess: try becoming a GM with math models. Its concepts are too abstract, like economy and people, for humans to grasp that way, even if engines show it’s possible
October 20, 2025 at 11:28 AM
Math is super helpful but a limited tool. What matters is if research is insightful, not how mathy it is.
Think of chess: try becoming a GM with math models. Its concepts are too abstract, like economy and people, for humans to grasp that way, even if engines show it’s possible
Think of chess: try becoming a GM with math models. Its concepts are too abstract, like economy and people, for humans to grasp that way, even if engines show it’s possible
If we compare the evolution of potential math engines to that of chess engines, it feels as though we're still in the early 1990s or even earlier. Yet the path from here to a Kasparov-Deep Blue moment appears open, so we might get there faster. I'm cautiously optimistic
1/2 Major AI systems have improved with symbolic search over the years, but they’re still not at the level of a “math engine” especially in terms of reliability. If & when that happens, such an engine could massively accelerate progress in the mathematical sciences including Econ & CS theory
October 18, 2025 at 2:15 PM
If we compare the evolution of potential math engines to that of chess engines, it feels as though we're still in the early 1990s or even earlier. Yet the path from here to a Kasparov-Deep Blue moment appears open, so we might get there faster. I'm cautiously optimistic
1/2 Major AI systems have improved with symbolic search over the years, but they’re still not at the level of a “math engine” especially in terms of reliability. If & when that happens, such an engine could massively accelerate progress in the mathematical sciences including Econ & CS theory
October 18, 2025 at 2:03 PM
1/2 Major AI systems have improved with symbolic search over the years, but they’re still not at the level of a “math engine” especially in terms of reliability. If & when that happens, such an engine could massively accelerate progress in the mathematical sciences including Econ & CS theory
I've seen, read, and met smart physicists who think social science is easy, so they can quickly make a groundbreaking contribution by 'proving' that the field has been wrong. The smarter ones engage and collaborate with social scientists, and actually make great contributions
October 17, 2025 at 8:44 AM
I've seen, read, and met smart physicists who think social science is easy, so they can quickly make a groundbreaking contribution by 'proving' that the field has been wrong. The smarter ones engage and collaborate with social scientists, and actually make great contributions