One of the reasons for creating this site was to address a need coming from artificial intelligence developers and project architects who wished to use their skills in the fight against climate change. This is not an easy pursuit, and I laud those individuals who have chosen this path to travel. For example, a person recently left Google and a six-figure job as a software engineer to trudge the muddy path toward climate solutions, a path that is muddy only because no one has previously paved the way forward with a solid plan of action.
Yoshua Bengio, Andrew Ng, and a number of other prominent researchers in the AI field put together a paper recently titled “Tackling Climate Change with Machine Learning.” They were very comprehensive in the ways machine learning (a subset of AI) could be put to use in mitigating the impacts of climate change. They addressed how machine learning could help in all these areas: electricity, transportation, buildings and cities, industry, farms and forests, CO2 removal, climate prediction, societal impacts, solar geoengineering, individual actions, collective decisions, education, and finance. Quite comprehensive indeed, but there is one important item that is missing.
Going back to the muddy path, we see that there is no structured plan forward, to make all the wonderful things highlighted in the paper actually happen with a well defined budget and timeframe. No solid statement of work exists, no work breakdown structure, no people identified who are responsible for tasks, no project plan. If there was ever any key item that artificial intelligence could help with in the fight against climate change, it is coming up with that project plan.
Engineers, here’s your problem. Do your magic and solve this one through natural language processing, geospatial analytics, whatever you think would be useful. This is the #1 problem that your skills could help solve that would have the greatest impact on climate change – simply formulating a plan for the rest of us to coordinate our activities.
Artificial intelligence these days manages to yield great insights despite the problems of bias inherent in training data and the amount of energy expenditure per insight gained. In trying to find technological solutions to climate change, one must be mindful of the carbon footprint of one’s technology search. However, it should not take much energy to use AI to develop a project plan, and the findings would be well worth the effort and time spent.