
Ajay Agrawal
@professor_ajay
learning | exploring | creating
ID: 949686988508139522
https://agrawal.ca/ 06-01-2018 16:59:57
1,1K Tweet
7,7K Followers
231 Following






Even the most sophisticated, modern AI systems are still just computational statistics that generate predictions. Rotman School Creative Destruction Lab (CDL)

Nice response Chris Hadfield

Where to find exceptional talent? The suburbs of #Toronto. tylercowen is spilling our secret! "The suburbs of Toronto are one of the world’s most neglected talent areas. Cities such as Mississauga or Brampton are now quite familiar to me, because so many Emergent Ventures


Privacy preservation is key to scaling AI. Richard Sutton 𝐒𝐞𝐧𝐝𝐡𝐢𝐥 𝐌𝐮𝐥𝐥𝐚𝐢𝐧𝐚𝐭𝐡𝐚𝐧 Ghost Bureau Content Services insights.intrepidgp.com/p/data-privacy…


In this application, AI appears to compress the productivity distribution (will likely reduce income distribution for workers in this domain) Rotman School Creative Destruction Lab (CDL)

AI => 15% productivity gain in radiology so far 9 years after... youtube.com/watch?v=2HMPRX… Ben Schamisso Northwestern Northwestern Radiology Northwestern Feinberg School of Medicine Rotman School University of Toronto

Productivity lift from AI: based on ~24,000 radiology reports [metrics: report completion times and clinical accuracy] Ben Schamisso Northwestern Northwestern Radiology Northwestern Feinberg School of Medicine



Developers who spend more hours on code-related activities per day were disproportionately faster with AI => AI does not benefit everyone equally => that's neither good nor bad, but it's important for anticipating labor market impact Elise Paradis 🦾💪🏿💪🏾💪🏽💪🏼💪🏻💪 Daye Nam Andrew Macvean Rotman School

I'm surprised this story is not receiving more attention. It's one of the best examples of what my coauthors, Joshua Gans and Avi Goldfarb, and I refer to as the difference between point solutions and system solutions in our book Power and Prediction. General Catalyst (GC) is

