How AI can impact the Global South

Thanks to Stanford Institute for Human-Centered Artificial Intelligence (HAI) & The Asia Foundation for your invitation last month to be a speaker at my alma mater Stanford University alongside leaders from Microsoft & the African Union Development Agency. 

Sitting in an intentionally curated room with fellow Tech executives from OpenAI, TikTok, NVIDIA etc, we discussed how AI can truly impact the Global South (developing countries) & more importantly the current limitations. 

The best part of the event - knowing some people in the world truly care about the welfare of the have-nots. One delegate even flew all the way from Nepal (over 30 hrs in transit) to stay only for a matter of hours and self-funds his research on this topic. 

Our sharing was under Chatham House Rules, but here are some personal reflections: 

 1. Whilst in developed countries we debate about whether generative AI technology such as chatGPT will replace white collar jobs, the developing world is looking to AI for solving basic everyday needs such as safety and obtaining a first time loan from a bank

 2. In the Global South, countries still struggle with instability of power grids (which prohibits their ability to benefit from AI innovation via online apps), the high costs of high bandwidth networking equipment, data gaps for local languages, lack of high quality digitized text for LLM training, and the environmental impact of climate change caused by manufacturing of more hardware & maintenance of data centers that are crucial to AI innovation

 3. For instance, in Malaysia’s affluent capital city Kuala Lumpur citizens can enjoy up to 800MB/second connection whereas in Sarawak broadband speed is only ~1-4MB/second. In Myanmar women are 39% less likely to use the internet due to affordability & literacy. Whereas in Singapore (which is only several hrs away by flight) > 99% of its population have access to high speed internet

 4. In order to bridge inequality, the Global South must fine tune their local data and develop astute methods for the application of LLM models to their proprietary data with a goal to improve the standard of living for the poorest in society

 5. AI can be used to upgrade standards of living by access to applications such as real time translation of languages, creation of credit proxies and democratizing education. Countries & organizations that can bridge the gaps to provide necessary infrastructure can make significant impact on these societies

 6. Like everything else in life, without deliberation the gaps between the haves & have-nots will only widen, we must intervene to provide equitable access for the Global South

hashtag#techforgood

Tech For Good Institute

For more research by Tech For Good Institute, visit: https://lnkd.in/gAZkMibG