Deena Mousa

China's rise as an innovation hub didn't spread via capital. Rather, entrepreneurs in emerging markets copied business models built under similar constraints New episode of my series with @vox_dev and @oliverhanney on AI in development with @realjoshlerner - we talk diffusion

This isn't an issue w @MattBruenig's piece specifically; it's intentionally US-centric. But when we talk about labor market impacts, we usually mean 'for US white collar workers', and I am much more concerned about people in countries that won't capture as much value from AI

The piece ends by saying that if the US becomes more inegalitarian, it won't be AI's fault but rather the policy choices we make. That's probably true for the US, but it is not really true for the world as a whole

Great piece in @TheArgumentMag about how to deal with income shocks due to AI labor market effects But one (big) problem: these solutions assume the country redistributing will have captured some gains from AI advancement. This is true for the US, but not for most of the world

Reading through the Anthropic survey answers coming from LMICs, I'm struck by how many of them are hopeful about LLMs allowing them to improve their individual economic circumstances -- I wonder how often this actually happens? And what the most common mechanisms are?

@vox_dev @oliverhanney ideasindevelopment.substack.com/p/raghuram-raj…

Are developing economies ready for the impact of AI on the service exports sector? On the latest episode of our @vox_dev series, @oliverhanney and I speak with Raghuram Rajan about India's growth prospects

China's rise as an innovation hub didn't spread via capital. Rather, entrepreneurs in emerging markets copied business models built under similar constraints New episode of my series with @vox_dev and @oliverhanney on AI in development with @realjoshlerner - we talk diffusion

What happens to a growth model built on services in the AI age? Raghuram Rajan joined @oliverhanney & @deenamousa @coeff_giving on Ideas in Development to discuss how India's economy grew through services exports & whether that model is resilient to AI: youtube.com/watch?v=blmDju…

In a new episode of my series with @vox_dev and @oliverhanney on AI in economic development, we discuss the many frictions of technological diffusion with @realjoshlerner voxdev.org/topic/technolo…

In 1925, hybrid corn came online in the US. It yielded ~20% more/acre, but by 1933 it was on just 0.1% of US corn acreage. It took another drought in 1936 to move Iowa farmers, Europe didn't follow until the 50s, and global adoption is still patchy. Could AI look the same?

Narrative violation - WFH doesn't make people lonelier? (h/t @notanastronomer)

@abdus1801 @venkmurthy Important to note the difference in the top row specifically is not statistically significant!

@deenamousa @venkmurthy According to first row, The Doctors performed poorer even with AI assistance than when they did their job independently??? How is the performance poorer even after AI assistance?

I recently got the chance to speak with radiology residents @UMassChen about AI's impact on their field This slide particularly resonated: a familiar pattern they've seen with AI assistive tools is having to nudge MDs to trust them *less*

* @UMassChan

A Kanchipuram Sari is not mere fabric. It is a language with grammar, a financial instrument you can wear, a biological object dyed with chemistry, & a cultural archive woven by history & hand. For over a millennium, it functioned as liquid wealth. In desperate times, the silk could be burned away, leaving behind real silver & gold. A garment that protected its wearer, even in fire. Today, that loom meets AI, microbes, & cryptography — systems trying to relearn lost grammar, grow colour without poisoning land, & turn trust from promise into proof. What happens when a thousand-year-old loom meets 21st century’s thinking machines? Issue 3: Kanchipuram Saris & Thinking Machines, written by Nivedita, designed & built by @AltCarbonIndia altermag.com/articles/kanch…