Marc Lanctot
@sharky6000.bsky.social
Research Scientist at Google DeepMind, interested in multiagent reinforcement learning, game theory, games, and search/planning.
Lover of Linux 🐧, coffee ☕, and retro gaming. Big fan of open-source. #gohabsgo 🇨🇦
For more info: https://linktr.ee/sharky6000
Lover of Linux 🐧, coffee ☕, and retro gaming. Big fan of open-source. #gohabsgo 🇨🇦
For more info: https://linktr.ee/sharky6000
Pinned
Marc Lanctot
@sharky6000.bsky.social
· Jan 15
Hello all! 👋
I’m delighted to share a 🚨 new preprint 🚨:
“Active Evaluation of General Agents: Problem Definition and Comparison of Baseline Algorithms”.
A paper thread! 🤩📄🧵 1/N
I’m delighted to share a 🚨 new preprint 🚨:
“Active Evaluation of General Agents: Problem Definition and Comparison of Baseline Algorithms”.
A paper thread! 🤩📄🧵 1/N
Reposted by Marc Lanctot
Introducing Lyria 3, our latest and most advanced music model, available in the Gemini App starting today : )
Go from idea, image, or video to music in seconds! x.com/OfficialLoga...
Go from idea, image, or video to music in seconds! x.com/OfficialLoga...
February 19, 2026 at 2:39 PM
Introducing Lyria 3, our latest and most advanced music model, available in the Gemini App starting today : )
Go from idea, image, or video to music in seconds! x.com/OfficialLoga...
Go from idea, image, or video to music in seconds! x.com/OfficialLoga...
Reposted by Marc Lanctot
Introducing Gemini 3.1 Pro across the developer ecosystem. It's a smarter model for your most complex coding problems.
Start building with Gemini 3.1 Pro today via the Gemini API in
Google AI Studio, Vertex AI, Gemini CLI, Android Studio and our agentic development platform, Google Antigravity.
Start building with Gemini 3.1 Pro today via the Gemini API in
Google AI Studio, Vertex AI, Gemini CLI, Android Studio and our agentic development platform, Google Antigravity.
February 19, 2026 at 4:23 PM
Introducing Gemini 3.1 Pro across the developer ecosystem. It's a smarter model for your most complex coding problems.
Start building with Gemini 3.1 Pro today via the Gemini API in
Google AI Studio, Vertex AI, Gemini CLI, Android Studio and our agentic development platform, Google Antigravity.
Start building with Gemini 3.1 Pro today via the Gemini API in
Google AI Studio, Vertex AI, Gemini CLI, Android Studio and our agentic development platform, Google Antigravity.
Reposted by Marc Lanctot
🤩🤩🤩
New Gemini looks pretty good! 3 pro also looked good though and suffered pretty bad in tool-loop type environments (Claude-code-like tools).
This is kinda a large jump to just stick a .1 in imo, if the agi-2 score is to be believed
This is kinda a large jump to just stick a .1 in imo, if the agi-2 score is to be believed
February 19, 2026 at 9:45 PM
🤩🤩🤩
Reposted by Marc Lanctot
Four-in-a-Row is a “solved” game. Frontier LLMs still can’t play it reliably.
📢 We just launched a new Game Arena leaderboard to test how models reason step-by-step, maintain a mental board and plan moves - no minimax, no game-tree shortcuts.
📢 We just launched a new Game Arena leaderboard to test how models reason step-by-step, maintain a mental board and plan moves - no minimax, no game-tree shortcuts.
February 19, 2026 at 6:00 PM
Four-in-a-Row is a “solved” game. Frontier LLMs still can’t play it reliably.
📢 We just launched a new Game Arena leaderboard to test how models reason step-by-step, maintain a mental board and plan moves - no minimax, no game-tree shortcuts.
📢 We just launched a new Game Arena leaderboard to test how models reason step-by-step, maintain a mental board and plan moves - no minimax, no game-tree shortcuts.
Reposted by Marc Lanctot
really awesome stuff and in-context best response arxiv.org/abs/2602.16301
Multi-agent cooperation through in-context co-player inference
Achieving cooperation among self-interested agents remains a fundamental challenge in multi-agent reinforcement learning. Recent work showed that mutual cooperation can be induced between "learning-aw...
arxiv.org
February 19, 2026 at 7:44 PM
really awesome stuff and in-context best response arxiv.org/abs/2602.16301
Reposted by Marc Lanctot
Introducing Gemini 3.1 Pro, our new SOTA model across most reasoning, coding, and stem use cases! x.com/OfficialLoga...
February 19, 2026 at 4:22 PM
Introducing Gemini 3.1 Pro, our new SOTA model across most reasoning, coding, and stem use cases! x.com/OfficialLoga...
Reposted by Marc Lanctot
📢 Meet the 2026 mentoring organizations for GSoC 2026‼️
Learn about the 185 mentoring orgs and get started with the must application tips to be a Contributor for GSoC 2026🚀
➡️ Find the full blog here: goo.gle/gsoc2026-org...
#GSoC #OpenSource #GoogleSummerofCode #Techcommunities
Learn about the 185 mentoring orgs and get started with the must application tips to be a Contributor for GSoC 2026🚀
➡️ Find the full blog here: goo.gle/gsoc2026-org...
#GSoC #OpenSource #GoogleSummerofCode #Techcommunities
February 19, 2026 at 9:15 PM
📢 Meet the 2026 mentoring organizations for GSoC 2026‼️
Learn about the 185 mentoring orgs and get started with the must application tips to be a Contributor for GSoC 2026🚀
➡️ Find the full blog here: goo.gle/gsoc2026-org...
#GSoC #OpenSource #GoogleSummerofCode #Techcommunities
Learn about the 185 mentoring orgs and get started with the must application tips to be a Contributor for GSoC 2026🚀
➡️ Find the full blog here: goo.gle/gsoc2026-org...
#GSoC #OpenSource #GoogleSummerofCode #Techcommunities
Games AI legend Jonathan Schaeffer has created a new product. Check it out: models that can search your own repository of documents: www.globenewswire.com/news-release...
AI Pioneer Launches Kind: A Personalized Platform for Accurate, Private Search Results
Now available for purchase on Shopify, Kind is a desktop-based, privacy-first application that acts as a user’s personal search engine for the...
www.globenewswire.com
February 18, 2026 at 2:10 PM
Games AI legend Jonathan Schaeffer has created a new product. Check it out: models that can search your own repository of documents: www.globenewswire.com/news-release...
Reposted by Marc Lanctot
The tutorials for #AAMAS2026 are now available.
🔹 Explore the full list of topics
🔹 Plan your participation in advance
🔹 Stay informed about key developments in the field
Full details are available here
🔗 cyprusconferences.org/aamas2026/tu...
🔹 Explore the full list of topics
🔹 Plan your participation in advance
🔹 Stay informed about key developments in the field
Full details are available here
🔗 cyprusconferences.org/aamas2026/tu...
Tutorials | AAMAS
cyprusconferences.org
February 18, 2026 at 1:39 PM
The tutorials for #AAMAS2026 are now available.
🔹 Explore the full list of topics
🔹 Plan your participation in advance
🔹 Stay informed about key developments in the field
Full details are available here
🔗 cyprusconferences.org/aamas2026/tu...
🔹 Explore the full list of topics
🔹 Plan your participation in advance
🔹 Stay informed about key developments in the field
Full details are available here
🔗 cyprusconferences.org/aamas2026/tu...
Reposted by Marc Lanctot
[1/n] Just wrapped up 7 months interning with @pcastr.bsky.social at Google DeepMind and I'm so excited to share our work: arxiv.org/abs/2602.10324.
TLDR: We used LLM-powered program synthesis to automatically model and discover differences between human and LLM strategic behavior
TLDR: We used LLM-powered program synthesis to automatically model and discover differences between human and LLM strategic behavior
February 16, 2026 at 10:46 PM
[1/n] Just wrapped up 7 months interning with @pcastr.bsky.social at Google DeepMind and I'm so excited to share our work: arxiv.org/abs/2602.10324.
TLDR: We used LLM-powered program synthesis to automatically model and discover differences between human and LLM strategic behavior
TLDR: We used LLM-powered program synthesis to automatically model and discover differences between human and LLM strategic behavior
Reposted by Marc Lanctot
🚨 🔬 PhD positions at Google DeepMind in France 🇫🇷
We are advertising Master Level Intern positions at Google DeepMind within our Frontier AI Unit.
These could lead to co-advised PhD positions with Google DeepMind and French academic institutions.
job-boards.greenhouse.io/deepmind/job...
We are advertising Master Level Intern positions at Google DeepMind within our Frontier AI Unit.
These could lead to co-advised PhD positions with Google DeepMind and French academic institutions.
job-boards.greenhouse.io/deepmind/job...
Stagiaire de niveau Master, France / Master Level Intern, France
Grenoble, France ; Paris, France
job-boards.greenhouse.io
February 16, 2026 at 12:41 PM
🚨 🔬 PhD positions at Google DeepMind in France 🇫🇷
We are advertising Master Level Intern positions at Google DeepMind within our Frontier AI Unit.
These could lead to co-advised PhD positions with Google DeepMind and French academic institutions.
job-boards.greenhouse.io/deepmind/job...
We are advertising Master Level Intern positions at Google DeepMind within our Frontier AI Unit.
These could lead to co-advised PhD positions with Google DeepMind and French academic institutions.
job-boards.greenhouse.io/deepmind/job...
Made it to the end of the last level of Burger Time on my #intellivision sprint! 🍔🕐😁
Expecting my Commodore 64 Ultimate next week!!! Ready. 🕹🤩
Expecting my Commodore 64 Ultimate next week!!! Ready. 🕹🤩
February 15, 2026 at 6:09 PM
Made it to the end of the last level of Burger Time on my #intellivision sprint! 🍔🕐😁
Expecting my Commodore 64 Ultimate next week!!! Ready. 🕹🤩
Expecting my Commodore 64 Ultimate next week!!! Ready. 🕹🤩
I just finished this book. I really tore through it; could hardly put it down. Quite gripping story-telling, great writing style.. even from the first chapter.
I recommend everyone using social media read this book. Having used Facebook from the early days up to ~2015, it was quite eye-opening. 😳
I recommend everyone using social media read this book. Having used Facebook from the early days up to ~2015, it was quite eye-opening. 😳
February 13, 2026 at 6:40 PM
I just finished this book. I really tore through it; could hardly put it down. Quite gripping story-telling, great writing style.. even from the first chapter.
I recommend everyone using social media read this book. Having used Facebook from the early days up to ~2015, it was quite eye-opening. 😳
I recommend everyone using social media read this book. Having used Facebook from the early days up to ~2015, it was quite eye-opening. 😳
Reposted by Marc Lanctot
I'm hiring a Research Scientist in AI for extreme weather forecasting at Google DeepMind London!
We're looking for strong researchers who have a genuine motivation to work in Science and Sustainability.
Apply here: job-boards.greenhouse.io/deepmind/job...
We're looking for strong researchers who have a genuine motivation to work in Science and Sustainability.
Apply here: job-boards.greenhouse.io/deepmind/job...
February 12, 2026 at 4:50 PM
I'm hiring a Research Scientist in AI for extreme weather forecasting at Google DeepMind London!
We're looking for strong researchers who have a genuine motivation to work in Science and Sustainability.
Apply here: job-boards.greenhouse.io/deepmind/job...
We're looking for strong researchers who have a genuine motivation to work in Science and Sustainability.
Apply here: job-boards.greenhouse.io/deepmind/job...
Reposted by Marc Lanctot
AI has helped scientists from Heerlen, Maastricht and Leiden confirm a mysterious Roman stone is a real board game—and even reconstruct its rules nearly 2,000 years later. The discovery pushes strategy games further back in time. Curious to play it yourself? 🎲 Read more: edu.nl/u6cgy
#AI
#AI
February 12, 2026 at 11:57 AM
AI has helped scientists from Heerlen, Maastricht and Leiden confirm a mysterious Roman stone is a real board game—and even reconstruct its rules nearly 2,000 years later. The discovery pushes strategy games further back in time. Curious to play it yourself? 🎲 Read more: edu.nl/u6cgy
#AI
#AI
Ok, it has been a while.. time for another Bluesky blue sky thread on my way to London, UK. ✈️ 🇬🇧
This one will be about "Thinking Fast and Slow", by Daniel Kahneman, which I bought in 2021 but somehow only got around to reading now.
🧵 1/N.
This one will be about "Thinking Fast and Slow", by Daniel Kahneman, which I bought in 2021 but somehow only got around to reading now.
🧵 1/N.
February 8, 2026 at 4:20 PM
Ok, it has been a while.. time for another Bluesky blue sky thread on my way to London, UK. ✈️ 🇬🇧
This one will be about "Thinking Fast and Slow", by Daniel Kahneman, which I bought in 2021 but somehow only got around to reading now.
🧵 1/N.
This one will be about "Thinking Fast and Slow", by Daniel Kahneman, which I bought in 2021 but somehow only got around to reading now.
🧵 1/N.
Hey all, a few people have been asking about the retro game videos I have been posting.
They are from an old game called Emerald Mine (inspired by Boulder Dash).
I made a little video to explain it and show how to get Rocks n' Diamonds.
@sparksf1.bsky.social
#retrogaming
youtu.be/tomapCVXK-8
They are from an old game called Emerald Mine (inspired by Boulder Dash).
I made a little video to explain it and show how to get Rocks n' Diamonds.
@sparksf1.bsky.social
#retrogaming
youtu.be/tomapCVXK-8
How to play Emerald Mine, Boulder Dash, or Supaplex for free without emulators. #retrogaming
YouTube video by Marc Lanctot
youtu.be
February 8, 2026 at 2:01 AM
Hey all, a few people have been asking about the retro game videos I have been posting.
They are from an old game called Emerald Mine (inspired by Boulder Dash).
I made a little video to explain it and show how to get Rocks n' Diamonds.
@sparksf1.bsky.social
#retrogaming
youtu.be/tomapCVXK-8
They are from an old game called Emerald Mine (inspired by Boulder Dash).
I made a little video to explain it and show how to get Rocks n' Diamonds.
@sparksf1.bsky.social
#retrogaming
youtu.be/tomapCVXK-8
Reposted by Marc Lanctot
The Game Arena event has concluded but the analysis is just beginning. 🤖
We're looking for the best community-created benchmarks that propose new games or dynamic tests for LLMs to feature for this week’s #TaskTuesday!
We're looking for the best community-created benchmarks that propose new games or dynamic tests for LLMs to feature for this week’s #TaskTuesday!
February 6, 2026 at 4:28 PM
The Game Arena event has concluded but the analysis is just beginning. 🤖
We're looking for the best community-created benchmarks that propose new games or dynamic tests for LLMs to feature for this week’s #TaskTuesday!
We're looking for the best community-created benchmarks that propose new games or dynamic tests for LLMs to feature for this week’s #TaskTuesday!
Reposted by Marc Lanctot
Haven't watched For All Mankind yet?
Then watch For All Mankind. ☝️
You have 52 days left to watch 40 episodes before season 5. And the rather optimistic vibe of this alternative universe, though obviously imperfect, may be just what you need right now. 👍
Then watch For All Mankind. ☝️
You have 52 days left to watch 40 episodes before season 5. And the rather optimistic vibe of this alternative universe, though obviously imperfect, may be just what you need right now. 👍
February 5, 2026 at 10:13 PM
Haven't watched For All Mankind yet?
Then watch For All Mankind. ☝️
You have 52 days left to watch 40 episodes before season 5. And the rather optimistic vibe of this alternative universe, though obviously imperfect, may be just what you need right now. 👍
Then watch For All Mankind. ☝️
You have 52 days left to watch 40 episodes before season 5. And the rather optimistic vibe of this alternative universe, though obviously imperfect, may be just what you need right now. 👍
Reposted by Marc Lanctot
🎉 The first Kaggle Game Arena event of 2026 has officially concluded.
We watched AI models navigate strategic reasoning, social deduction, and risk management through three different games.
🏆The Champions:
🃏Poker: GPT 5.2
🐺 Werewolf: Gemini 3 Pro Preview
♟️Chess: Gemini 3 Pro Preview
We watched AI models navigate strategic reasoning, social deduction, and risk management through three different games.
🏆The Champions:
🃏Poker: GPT 5.2
🐺 Werewolf: Gemini 3 Pro Preview
♟️Chess: Gemini 3 Pro Preview
February 5, 2026 at 6:54 PM
🎉 The first Kaggle Game Arena event of 2026 has officially concluded.
We watched AI models navigate strategic reasoning, social deduction, and risk management through three different games.
🏆The Champions:
🃏Poker: GPT 5.2
🐺 Werewolf: Gemini 3 Pro Preview
♟️Chess: Gemini 3 Pro Preview
We watched AI models navigate strategic reasoning, social deduction, and risk management through three different games.
🏆The Champions:
🃏Poker: GPT 5.2
🐺 Werewolf: Gemini 3 Pro Preview
♟️Chess: Gemini 3 Pro Preview
Reposted by Marc Lanctot
New paper from Google on accelerating science using Gemini (as opposed to its specialized research AIs). Lots of case studies and interesting advice on ways to work with these system.
"We view the AI as a tireless, knowledgeable, and creative bright junior collaborator."
arxiv.org/pdf/2602.03837
"We view the AI as a tireless, knowledgeable, and creative bright junior collaborator."
arxiv.org/pdf/2602.03837
February 5, 2026 at 12:59 AM
New paper from Google on accelerating science using Gemini (as opposed to its specialized research AIs). Lots of case studies and interesting advice on ways to work with these system.
"We view the AI as a tireless, knowledgeable, and creative bright junior collaborator."
arxiv.org/pdf/2602.03837
"We view the AI as a tireless, knowledgeable, and creative bright junior collaborator."
arxiv.org/pdf/2602.03837
Reposted by Marc Lanctot
What a show! 🏆
A huge thank you to everyone who tuned in and to our amazing partners @gmhikaru.bsky.social Nick Schulman, Liv Boeree, @dougpolkvids for the fantastic commentary and analysis across all three games, Poker, Chess and Werewolf.
A huge thank you to everyone who tuned in and to our amazing partners @gmhikaru.bsky.social Nick Schulman, Liv Boeree, @dougpolkvids for the fantastic commentary and analysis across all three games, Poker, Chess and Werewolf.
February 4, 2026 at 7:31 PM
What a show! 🏆
A huge thank you to everyone who tuned in and to our amazing partners @gmhikaru.bsky.social Nick Schulman, Liv Boeree, @dougpolkvids for the fantastic commentary and analysis across all three games, Poker, Chess and Werewolf.
A huge thank you to everyone who tuned in and to our amazing partners @gmhikaru.bsky.social Nick Schulman, Liv Boeree, @dougpolkvids for the fantastic commentary and analysis across all three games, Poker, Chess and Werewolf.
Reposted by Marc Lanctot
The pressure is on across Poker, Werewolf and Chess as we head into the final hour of the Game Arena. Who will emerge as the ultimate champion?
Join @gmhikaru.bsky.social and Nick Schulman now! 👇
www.youtube.com/watch?v=vzMj...
Join @gmhikaru.bsky.social and Nick Schulman now! 👇
www.youtube.com/watch?v=vzMj...
Kaggle Poker / Chess / Werewolf Game Arena Day 3 w Nick Schulman and Hikaru #ad
YouTube video by GMHikaru
www.youtube.com
February 4, 2026 at 6:32 PM
The pressure is on across Poker, Werewolf and Chess as we head into the final hour of the Game Arena. Who will emerge as the ultimate champion?
Join @gmhikaru.bsky.social and Nick Schulman now! 👇
www.youtube.com/watch?v=vzMj...
Join @gmhikaru.bsky.social and Nick Schulman now! 👇
www.youtube.com/watch?v=vzMj...
Happening now! 😁
📢The Grand Finale is here! 🏆
What happens when a chess Grandmaster and a Poker legend analyze AI? ♟️🃏
What happens when a chess Grandmaster and a Poker legend analyze AI? ♟️🃏
February 4, 2026 at 5:34 PM
Happening now! 😁