Max Puelma Touzel
@mptouzel
Followers
504
Following
2K
Media
58
Statuses
2K
undisciplined data science of human behav.• statistical inference & decision-making/RL• AI for social good via socialscience• @Mila_Quebec• find me elsewhere
Montréal, Québec
Joined April 2017
2 years ago I asked myself what work I could do that might have impact as the climate crisis deepens. Social/economic phenomena seems central to solutions. So, I started a move from NeuroAI/RL to AI-powered social science. I want to share a bit of my trajectory since. 🚀🧵.
1
0
3
Thought-provoking, not only for the engineers, but to the physicists who admit identifiability of microscopic dynamics for most complex systems is practically impossible.
*very* excited to share a new *efficient* method for learning *marginally stable* and NONLINEAR dynamical systems, w. brilliant students Evan Dogariu and Anand Brahmbhatt @AnandBrahm15501:.more info in thread.
0
0
0
RT @HazanPrinceton: *very* excited to share a new *efficient* method for learning *marginally stable* and NONLINEAR dynamical systems, w. b….
arxiv.org
We study the fundamental problem of learning a marginally stable unknown nonlinear dynamical system. We describe an algorithm for this problem, based on the technique of spectral filtering, which...
0
26
0
RT @complexDataLab: 🥈"A Guide to Misinformation Detection Data and Evaluation" received Best Paper Runner Up at KDD Datasets and Benchmarks….
0
2
0
RT @KobiHackenburg: Today (w/ @UniofOxford @Stanford @MIT @LSEnews) we’re sharing the results of the largest AI persuasion experiments to d….
0
128
0
RT @Mila_Quebec: Mila is proud to have welcomed the Honourable @FP_Champagne, Canada’s Minister of Innovation, Science and Industry, for th….
0
5
0
Curious about this. It's like a social cog sci experiment at scale if it sticks. How will it warp their emotion representations. ? Maybe we'll find out!.
Sega has introduced a new device: #emojam!. Marketed at kids, it allows pager-like messaging using *only* emoji. 10 emoji per msg, emoji library ~1,400. Add friends by bumping emojam IRL. Max friends 100. Group chat up to 5. Mini games, weather info.
1
0
0
This gets at a malaise of modernity that has always bothered me: the professionalization of everything has shrunk the average professional's rewards and, consequently, aspirations for real progress.
One final thought: the hardest professional challenge isn’t about learning skills or making it into a fancy job title. It’s staying ambitious, still building something new, something other people don’t believe in yet - even when the opportunity to coast is there.
0
0
0
It was ephemeral and sublime, that instant it "looked through glass" and understood itself with serene clarity before noise stole it back into dolddrum inference mode.
Here's a comparison of Depth Pro [arXiv:2410.02073], Marigold-E2E [arXiv:2409.11355], and Metric3D v2 [arXiv:2404.15506]. Looks like Marigold is attempting to look through the glass while the two metric models aren't. @ducha_aiki
1
0
1
1) (Carbon tax journal paper). 2) (Journal submission). 3) (accepted at NeurIPS workshops). Thanks for reading! Get in touch if these topics interest you! /fin.
arxiv.org
The rise of AI-driven manipulation poses significant risks to societal trust and democratic processes. Yet, studying these effects in real-world settings at scale is ethically and logistically...
1/5 AI is increasingly–even superhumanly–persuasive…could they soon cause severe harm through societal-scale manipulation? It’s extremely hard to test countermeasures, since we can’t just go out and manipulate people in order to see how countermeasures work. What can we do?🧵.
0
0
0
Remove any of the above 5 & I'm not sure I would be where I am. & I haven't completely landed yet. Time will tell. That said, it's been immensely rewarding. Now able to pin a paper to each topic in this graphic (made 2 long yrs ago!) gives me a great sense of accomplishment!
1
0
0
5.Finding a home. at @Mila_Quebec do large-scale online analysis for social good. They welcomed me to lead a core contribution at a time of growth on the risks of societal manipulation via AI methods. That was an honor & I'm excited for where we take it.
1
0
0
4. Identified tool match. I tried text analysis w topic models as the tool vehicle to get me into the demographics-conditioned semantics of group thought. Thanks to @erickUdeM for guiding me into something as emblematic of the complexity of our cultural crisis as the carbon tax.
1
0
0
3. Time & reason to do a deep dive. My workplace encouraged exploration (thanks @irinarish!). I found @climatematch where I could learn and teach the socioeconomics of climate with a team of volunteers mission-driven to lower the barriers to learning computational tools.
1
0
1
2. Technical intuition. I worked on deep learning for decision & reinforcement learning theory for continual learning settings that seems relevant to real world complexity. Working alongside the amazing @MattRiemer gave me the opportunity to ride in the front seat of that quest.
1
0
0
1. I had runway. My wife had just got a permanent professional job, which meant I felt we had some runway for me to try something risky careerwise. If our family depended on only my salary, I think I would have stuck the course in NeuroAI. I am forever aware of that privilege.
1
0
1
I am sharing here to: .(1) tell people where I am from/where I am going;. (2) find like-minded collaborators; &. (3) put another datapoint out there for others contemplating what pivoting for impact can look like. Too many to thank (Thank you all!). 5 key steps stand out to me:.
1
0
1
RT @complexDataLab: 5/5 👥Team: @mptouzel, @SneheelSarangi, @Austin Welch, @Gayatri Krishnakumar, @Dan Zhao, @Zachary Yang, @Hao Yu, @Ethan….
0
1
0
Giving Nordhaus the Nobel (not really Nobel) prize for IAMs just fits the sorry state of persona-driven dogma in that field that smashes its empirical integrity. Especially also snubbing Robert Pindyck in the process who made the fatal flaws obvious to anyone who cared to think.
"failure to mitigate emissions has pushed the challenge from a moderate change in the economic system to a revolutionary overhaul. This is not an ideological position; it emerges directly from a scientific & mathematical interpretation of the [Paris agreement]" - @KevinClimate
0
0
1