Rich countries are also aging into a permanent care shortage. This physical, relational work is among the hardest for AI to automate. Home health aides are the fastest-growing US job by a wide margin: +740k by 2034, nearly 3x the runner-up.

So why migration? Comparable workers (same age, schooling, experience) earn much more in the US than at home. If AI makes rich countries richer while that work stays human, that gap will only widen.

Selling clean energy runs into another issue: demand for green hydrogen is manufactured by regulation and hasn't manifested yet. ~90% of the world's announced 2030 supply has no financing behind it.X$

Data centers look like an opening, since the US is blocking them amid local opposition. But for poor countries they mostly aren't exports: 80%+ of African data already sits abroad, a hyperscale campus hires ~150 people, and the margins go to the operator.

Tourism only works if you have something enticing for visitors. Even then, there's a limitation on demand and how many people you can serve. Of the US sectors tracked since 1990, accommodation and food is nearly the only one where hours grew faster than output/worker.

Another is agriculture - people need food, and AI is not growing it. But demand for food is capped, so when productivity rises, workers leave farming. And the real money is in the branding and marketing; farmers capture about 6.6 cents of every chocolate dollar.

Export-led growth did 3 things: earned hard currency, employed huge numbers of people, and made those workers steadily more productive by doing the work. High revenues alone aren't enough if fewer and fewer people are employed by that industry, as I wrote in my last post.

NEW: If AI erodes paths to growth, like manufacturing and service exports, what can poor countries do? I went through commonly-recommended contenders and came out skeptical that any are a truly comparable growth engine. Is sending people to rich countries the best bet?

But the same traits that make a job possible to offshore also make it especially automatable. While there's no sign of disruption looking at overall employment, new jobs in these fields are declining.

Manufacturing employment used to peak when countries were already rich, at around $35,000 per capita for Britain, Sweden, Italy. But later industrializers are hitting the same peak at about $1,800.

Manufacturing as a path is already closing. The sector has gotten so competitive that the industry cannot create enough jobs. Bangladesh's garment exports tripled between 2010 and 2024, while employment in the sector stayed flat.

New: AI could keep poor countries poor by eliminating paths to economic growth. Stocks in Indian IT firms and Philippines' call centers have been selling off, driven by fears about AI automation. In an age of AI, will the world need any of what poor countries sell?

The lesson is that we ought to be *far* more ambitious. Maintaining a rat-free province sounds impossible and onerous, and if attempted elsewhere would likely garner complaints about wasted taxpayer funds. But it turns out to be *cheaper* than half-hearted reactive measures.

Today Alberta mans a rat sighting hotline. Of 875 reports last year, only 47 were rats; Albertans have never seen the creature, so they often mistake squirrels for one. Officials tracked a pet rat down via a license plate in a FB photo. Another was reported by an ex (they re-homed it to BC!)

They adopted Soviet-style war posters, hosted radio shows about the dangers of ignoring rats, and sent community educators to run town halls. One chewed on warfarin-treated oats while taking questions to assuage residents' concerns about the pesticide, so they'd stop throwing out treated traps.

Albertans - worried about disease and crops - called a state of emergency. They set up a 'Rat Control Zone' on the border, covered it in pesticides, and sent inspectors to set up watch. To rally enough people, some bureaucrats were reassigned from weed inspection to rat patrols.

The rats we're familiar with aren't native to North America. They hitched rides on ships with Europeans, and, by the late 1940s, they'd crossed Saskatchewan and were approaching Alberta's border at 24km/year.

Rats are virtually everywhere humans live. Except for Alberta, Canada. Most residents there have never even seen one. How did they eliminate a pest we take as a fact of life? By waging war.

We've gotten used to technology being democratizing but, unfortunately, it seems unlikely AI will be a leapfrog miracle story. New post out now on why: newsletter.deenamousa.com/p/ai-wont-b...

New post: We don't know why Malawi is so poor newsletter.deenamousa.com/p/we-dont-k...

New: Big technological advances in health, agriculture, and connectivity have not moved Malawi out of subsistence — and there's no war or state collapse to blame. Without knowing why Malawi is poor, you can't predict what AI will do for growth in low-income countries.

I recently got the chance to speak with radiology residents @umasschan.bsky.social 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*

Looking at workers in industrial jobs exposed to automation, those who worked with robots directly were more likely to believe both that (a) their job was automatable and (b) robots would create new job opportunities for them.

College students would need to be paid $59 (on average) to deactivate TikTok for a month, but would *pay* $28 for it to be deactivated for them and their social circles for the same amount of time.

The “AI bubble” may be less catastrophic than it sounds. Even if it pops, it’s leaving behind real infrastructure, much like past bubbles did. My take for @TheMorningNews.org 2025 year-in-review post:

New post: ML trained on phone usage data can help identify households in need during crises. Not yet finalized your end of year giving? I've teamed up with GiveDirectly along with a few other Substackers, and I’ll be matching the first $500 in donations through the link in post.

There are several devices in clinical use today, like PainChek, a smartphone app that scans the facial expressions of people who have dementia and uses AI to output an expected pain score to inform their care. They also record data for the patient and facility over time.

Just out in @technologyreview.com: I look at the companies using AI to measure how much pain patients are in based on everything from involuntary facial movements, to heart rate, to peripheral temperature changes. Will this oust the classic self-reported 1-10 scale?

A new paper in the Lancet finds that doctors got *worse* at finding precancerous growths during colonoscopies on their own if they had just spent three months using an AI assistive tool Worrisome sign that deskilling may happen a lot faster than we'd expect.

We also don't fully understand why many people experience phantom limb pain, and many others don't x.com/deenamousa/...