Toward Universally Benevolent, Ethical AI

.21.01.2021

By the end of this decade, artificial intelligence governed by ethical AI principles will direct international commerce. It will assist us in practicing law and medicine. It will guide human teachers in the classroom, allowing personalized education to ensure that all our children reap the promise and benefits of education. AI will be deeply and irrevocably involved in the production and distribution of our food supply and will be the reason we prevent future pandemics and control climate change.

Human intelligence—and its artificial counterpart in AI—will soon become irreversibly inseparable through mutual need and the relentless gravity of human progress. But the shape of this future and our path toward it are not yet set in stone. The default path of leaving AI’s development to the whims of Big Tech, as we did with software these past four decades, is a dangerous option. After all, AI is imminently more powerful and harder to control than software, which is why it is widely recognized for exacerbating inequality, increasing extremism, and sowing societal discord through disinformation. As such, the urgency to reframe our AI progress is real and warranted. If there is a way to do better with AI, we have a moral obligation to find it and commit to that change. It is also important that we proceed in a thoughtful manner with both regulation and innovation to realize our ethical AI future. We owe it not only to ourselves, but to posterity.

It is a time for reason, empathy, and visionary boldness to work synergistically to make AI a net benefit for humanity. This manifesto is intended as an educational tool about the subtleties of our transition from a dedicated software world to an ethical AI one, and also a blueprint of the pitfalls we will encounter along the way.