Uneven AI distribution…
Interesting new research from Microsoft on the growing developed vs developing world gaps when it comes to AI diffusion. According to Microsoft over 1.2B people use AI but the 4B who lack electricity, internet, or digital skills are being frozen out. AI adoption is 2x in the Global North vs the Global South. Half the web is English (spoken only by 5% natively). And so on.
Small infrastructure gaps appear, again, to becoming civilizational dividers. I say ‘again’ as the themes touched on by Microsoft are very similar to those covered in Jared Diamond’s seminal ‘Guns, Germs and Steel’. In that he argued that geographic advantages (climate, domesticable animals, navigable coastlines) led to early agricultural surpluses in Eurasia, which enabled a specialization, population density, technological development, and ultimately military dominance. Societies without these initial advantages fell increasingly behind. Microsoft appear to argue the same: geographic advantages (electricity infrastructure, internet connectivity, English-language dominance, GDP) are enabling early AI adoption, which will create compounding advantages in education, productivity, and economic development. Societies without these “building blocks” risk falling increasingly behind.
Are we watching the creation of a new technological caste system, where AI-enabled societies pull so far ahead that catching up becomes structurally impossible, just as Diamond argued happened with agriculture, writing, and industrialization?
Maybe.
Diamond has been criticised for downplaying the role of human agency and policy choices. Microsoft’s report reinforces this criticism, by showing that human decisions (like Singapore’s 1980s digital investments) can overcome initial disadvantages.
Whilst infrastructure gaps create similar patterns of inequality, these are human-created systems that can be changed through investment and policy. Would be great to see more of investment in this direction, amongst all the multi-billion dollar investment news headlines we are seeing these days.
-
Global AI Adoption in 2025 – A Widening Digital Divide
