A Machine Learning Algorithm Goes Over Data From Thousands Of Articles To Reveal That 85% Of The World's Population Is Affected By Climate Change

A new AI (Artificial intelligence) analysis reveals that man-made climate change is already affecting around 85% of the world’s population in one way or another.

A publication in the Nature Climate Change journal, made by a total of 19 researchers and collaborators, adds more fuel to the climate change debate and adds more evidence to the stance that climate change isn’t speedily heading our way: it’s already here. To be fair, calling the entire exchange a debate seems to posit that either side could be correct. That, frankly, is very incorrect. Climate change is a fact, not a hypothesis, and the fact that its effects are already running rampant has been put forth by many notable scientists and organizations across the world. The United Nations even went so far as to call a press conference, with the current General Secretary Antonio Guterres declaring that unless we were to change our pace, humanity’s time on this planet would be fraught with endless problems.

The research was conducted by utilizing a machine learning program called the BERT language model. Upon being trained by poring over more than 2,000 separate hand coded documents, the AI was then set loose amongst a slew of research papers between the years 1951 and 2018. BERT managed to make its way through over ten thousand documents covering climate change and its effects upon society. More than that, the AI also identified important details such as location and time, which were then mapped out by researchers in order to accurately determine the impact across the globe region by region. The conclusion that 10,000 papers led to? Climate change is affecting individuals. A lot of them. What a surprise.

Data analysis revealed that anthropogenic climate change already affects over 80% of the Earth's landmass, which accounts for 85% of the planet's population as well. While that's already very worrying, there's more bad news on the way. Literature on climate change was typically found to be aimed at first world countries, with research in third world countries being relatively sparse. Therefore, with BART having no way to compensate for the difference in data, there's a very good chance that the estimated 85% of our affected population is in and of itself an underestimation.

This very gap in data should call for more research to be conducted in developing countries, if only so the world can have more information about the extent of the damage.


Read next: Can AI Manage HR Better Than Actual Humans?
Previous Post Next Post