Karthik Tadepalli
@karthiktadepall
Followers
2K
Following
2K
Media
96
Statuses
882
i want the rest of the world to get to the moon. PhD @berkeleyecon
Berkeley, CA
Joined June 2017
In 1960, Brazil was receiving food aid from the US. Today, it is the world's bread basket. The key was creating a cutting-edge agriculture research institute, Embrapa – an icon of successful science policy that most people have never heard of. 🧵
41
725
3K
Are we ready for a world in which humans possess an ever-shrinking fraction of the knowledge on which science and technology operate? How much can AI really accelerate science if constrained so that humans have to understand that progress?
If the the predictions that AI will create minor scientific discoveries next year and major ones a couple years later come true, it is worth noting that we have no real mechanism in academia for accommodating, reviewing, processing, and disseminating a sudden increase in science.
0
0
2
I've started a second Substack, mostly focused on art and fiction. Stuff that is important to me but not what people know me for on Twitter. Today I posted the first short story I actually want people to read :)
1
0
5
20th century global development was defined by technology transfer agreements – formalized transfer of knowledge and machines from advanced economics to developing countries. But today, technology transfer agreements have vanished. My latest article explores why:
2
9
33
What were the causes of South Korea's export miracle? My new paper w @philippbarteska, @straightedge, @seung_econ explores a major overlooked factor: U.S. military procurement during the Vietnam War. We bring new evidence to the Q of how geopolitics shapes development. 🧵:
9
125
346
Me after I once asked Claude to explain Lacan like I'm 5
@TadeJanda @the_mel_jar Armed with a rudimentary understanding of Lacan, I’m about to become insufferable 😄
0
0
2
"be yourself" as the go-to cliche has always been wild to me. For ~all of history you were not supposed to be yourself! You were supposed to be your social role, no matter what you wanted. Telling people to live in harmony with themselves is groundbreaking new technology!
i don’t know anything but my experience of aging so far has been a series of sudden& humiliating realizations that every cliche made meaningless by constant exposure (“be yourself” etc) contains a completely overwhelming degree of truth that u can’t understand until you just do
0
1
28
Incredible how strong is people's desire for others to tell them who they really Are, so that they know how to Live
I'm obsessed with this subject—how our ability to assign clinical labels to every feature of our existence both helps and hurts us—and honestly, there's a modern Borges story to be written about a future DSM [the psychiatric diagnosis bible] that grows larger and larger until
1
1
4
I wrote a long blot post on the economics of AI, prompted by two great workshops (Windfall and NBER), and me leaving OpenAI for METR: TLDR: we driving in the fog. 1. There is no standard model of AI’s economic impact. Economists have been using a wide range of assumptions to
22
94
472
Accidentally said "opinion" instead of "prior" and they kicked me out of Berkeley
Accidentally said "hard" instead of "non-trivial" and they kicked me out of SF
1
0
28
I hear this claim all the time and I don't understand it! Capital is already allowed to substitute for labor in the models, AGI would simply be capital that can substitute for labor at all tasks. What's broken?
Any economic concept or model that takes as a given the ideas of "Labor" and "Capital" (as they're currently defined) is likely to break in the presence of Artificial General Intelligence
1
0
16
Nobody thinks about the economics of transformative AI as clearly as Phil Trammell. And now I don't have to keep paraphrasing his ideas, I can just point you to this great podcast with Epoch! https://t.co/MvpfnMdr6W
epoch.ai
Stanford economist Phil Trammell joins Epoch AI to explore AGI, growth, GDP limits, and what economic theory can tells us about the future of AI.
0
0
7
TWFE defenders rejoicing at this graph
A #DiD post after a long time but this one is worth bookmarking! This 2025 #APSR paper conducts a massive reanalysis of recently-published papers in key political science journals using different DiD methods: Causal Panel Analysis under Parallel Trends: Lessons from a Large
5
12
145
Good thread I think a related issue is that most students are not serious about their research. Most are somewhat achievement-motivated. But they aren't deeply trying to understand anything, and rhat shows in their actions/output
Frequent convo between me and PhD students Me: "After 400 hours of data analysis on industry X, have you talked to anyone actually working in industry X" Student (S): "No" Me: "Why not?" Student: "Where would I find someone in industry X willing to talk to me"
19
30
513
College dining hall food was so bad that it drove me to start cooking in an instant pot in my 50 sq ft dorm. Graduating to a real kitchen was heavenly. Now cooking is my great life joy
I would have a mental breakdown if I got to eat unlimited high quality sushi every day for all of undergrad and then had to start cooking my own food. Good luck to these kids
1
0
8
Since this is blowing up, check out this recent paper rigorously evaluating Embrapa's effects! It's also how I heard about Embrapa. https://t.co/WGZq18D1ea
Really excited to share a new WP, together with Ariel Akerman, Jacob Moscona and Karthik Sastry, about the role of public R&D in Brazil's agricultural revolution. ( https://t.co/64OX3AwLSL) A quick thread.
2
2
50
See the full story, featuring a reluctant founder press-ganged into service, cows coming into Sao Paulo as a publicity stunt, and how Embrapa's strategy was secretly a massive win for the Chicago school of economics.
6
12
235
Embrapa's success is iconic because "invest in R&D" is not part of the standard development playbook. And because it shows that directed research programs can achieve industrial policy goals (growth in a targeted sector). It is a model for any country with higher ambitions.
4
15
261
Lesson #3: respond to market demand. Embrapa focused its early efforts on expanding arable land because a big constraint on Brazil's ag was rising land prices. They also targeted higher price crops. This "induced innovation" strategy gave direction and avoided a fractured focus.
1
12
257