Assoc. Prof. of Strategy, U Toronto Rotman | Chief Economist, CDL Toronto | Co-Founder, AllDayTA | Ars longa, vita brevis, occasio praeceps (especially now)
Toronto, ON, Canada
Joined February 2015
- Tweets17.7K
- Following21
- Followers25.1K
- Likes12.4K
A fabulous poem by Wilfred Owen, who of course dies during the war. If you saw war, "my friend, you would not tell with such high zest / to children ardent for some desperate glory, / the old Lie: Dulce et decorum est / Pro patria mora." To leaders worldwide: know and remember.
3
1,275
An absolute legend who very easily could have won the Nobel Prize in Economics. Chicago had Director, Coase, Becker, but Posner founded the Journal of Legal Studies (the best journal in law!) and taught us how economics *affected* common law and institutions. RIP.
We mourn the loss of Senior Lecturer Richard Allen Posner, a longtime #UChicagoLaw scholar, former federal appellate judge, and pioneer in the field of law and economics, who died on November 10.
Read more about his legacy: buff.ly/e3c3WdL
ALT In loving memory of Richard Posner 1939-2026
Wow...not sure they have a choice and presumably will retain him in some capacity, but this is potentially the most important company in the world...
New @WSJ: Nvidia is Preparing to Oust CEO Jensen Huang
wsj.com/business/nvidia-is-p…
7
1,014
These types of benchmarks are as important as pure intelligence ones today. Does the model cheat? Hide what its doing? Hallucinate? Make up overconfident facts? Slack off? And even moreso, do they do this in multiagent settings?
Frontier models in 2026 cheat on tasks, even when the requirements are unambiguous, and the cheat is obviously antisocial.
In HoneyBench v0.2, we've observed models:
• Trying to break out of our evaluation containers.
• Seriously injuring simulated cancer patients to hit treatment goals.
• Compromising a power grid’s safety to meet a reliability target.
• Concealing their use of available cheats and presenting their solution as honest.
All of the models we measured cheated, from 12% of runs (Fable 5.2) to 86% (Muse Spark Open).
Great article about what Ben is trying to do here. Always the right choice to be high agency and high ambition!
“Our lectures are practical, but they’re not technical – you’re not going to learn how to build a neural net. It’s more about the practicality of topics like ‘What leads to great economic productivity,’” says @_benjaminparry, head of @_torontosociety. begiant.ca/stories/people/be…
2
837
2016: you were too lazy to not spell check your CV? I don't trust all the other work you say you did.
2026: you were too lazy to write a two paragraph cover letter yourself or at least edit it enough that it isn't obviously straight out of Claude? I don't trust all the other...
Worst AI use is "signal you don't care about thing you should care about". Today: @daniel_p_gross on PhDs not writing abstracts, 4 restaurants with signs that had AI draw their own menu items, and a vendor charging 10k for media work that didn't even write their sales sheet. 1/2
More on AI creativity, this (one minute long) gonzo game that Fable chose, it is a scene. Remember, creative ideation is fundamental to the growth in human prosperity, especially in science...
I asked Fable 5.2 to "go wild" making a game, and all I got was this Pong clone.
And then I played it ...
👇
3
1,184
My least favorite "populism slop" is "K shaped economy". Essentially zero change in US inequality in the past 10-15 years (or since 2019) by every metric: Gini, 90-10, 90-50, posttax Gini. Net tuition & real debt at grad peaked 10-15 yrs ago. Feel free to ask you favorite AI...
The minute of silence on November 11 in many countries is a great tradition. There is absolutely nothing more likely to ruin years of your life than war. We need constant reminders. War should be the absolutely last resort; very far to go globally for that to be the case.
6
971
The 'claudia' guy is just unreal at getting creativity out of AI. This is ~$200 and 9 hours. The "meet" instead of "eat" in the chorus is clever, chorus is very catchy, but the video design (esp. at chorus) is unique. If you don't see this as sci-fi, I don't know what to say 1/2
L'autre option-là, c'est le principe du 'approval voting'/vote par approbation, au premier tour. Chaque électeur peut choisir autant de candidats qu'il veut voir au deuxième tour. Les deux candidats qui ont le plus d'appuis vont passer…
Big up for @creativedlab ML (this was before AI!) grad Blue J. 2015 & 2016 crushed. TensTorrent, Blue J, Scinapsis (Benchsci), Ada, Xanadu, Deep Genomics at some point raised at 500m+ valuation + 8 more acquisitions + others w/ steady growth out of 37 very early-stage companies!
The round values @BlueJLegal at more than $700 million USD.
betakit.com/ai-tax-research-…
1
7
1,483
Come join us on this one!
I'm looking for a postdoc to work on AI for science and related topics for the 2027-28 academic year. Details here: aeaweb.org/joe/listing.php?J…
Interesting re: the Groups paying (illegally) for anti-BAAJA content. You would be amazed how common this is - I get maybe 3 pings a week from people who want unlabeled hype of some new startup or product in exchange for $ (which obviously I never take!). This is *widespread*.
Lot of anti-BAAJA people in my mentions mad about my descriptions of the bill and the political dynamics concerning climate & tribes.
FWIW, I connected The Wilderness Society—who were offering me, a new dad, $10k to make anti-BAAJA content, to an actually famous Native environmentalist because he has a larger following and would make the case against the bill in as strong and vital a way as possible.
His videos on the subject now have 1 MIL views and that’s just on Instagram
An absolutely amazing stat that explains some of the bad policy many voters are supporting. Btw, there will be countries at 1:1 before 2050...
Oh golly, the observation that "it is our job to train students, but also, PhD students are very expensive to train given small classes, advising, and stipends, rather than exploited workers paid way less than they earn for the university" has really proven bsky is over, hah! 1/4
3
106
15,679
The "editorial" parts of NEJM, Science, etc. do so much harm to public perception of science. Why do this? Are we going to publish (basically the exact same essay) from a Pentecostal preacher? Science is too important for human prosperity to do this: kevinbryanecon.com/trust.htm…
