Impromptu: Amplifying Our Humanity Through AI - AI can be used to help us analyze and understand large and complex data sets, which could help us better address issues like climate change, food insecurity, and poverty. Naturally this universal upgrade to humanity is not without complications. The same capabilities and services that can help an unbanked farmer in a developing country connect with new customers and sources of capital can also empower ransomware gangs, terrorist organizations, and racist conspiracy-mongers. And, because we human beings are calibrated to perceive the awful more acutely than the good, we often see further innova- tion not as a potential solution—aka progress—but as the cause of some (or most) of our pressing problems. We anticipate even greater problems down the road. GPT-4: From an evolutionary standpoint, it makes sense for us to be highly attuned to danger or threat. If you’re a caveman and you hear rustling in the bushes, it’s more important for you to be worried about a potential pred- ator than to be excited about a potential meal. Being on high alert for the bad things in life can help us survive. Additionally, I think the negative tends to stand out more to us because it’s often more unexpected or dis- ruptive. We’re used to things going relatively smoothly in our lives, so when something comes along that disrupts that, it’s jarring and grabs our attention. Finally, I think there’s an element of negativity bias that comes from our culture and media. We’re constantly bombarded with bad news and stories of tragedy, which can make it seem like the world is just full of awful things. This can condition us to be hyper-focused on the negative, even when there’s plenty of good in the world too. 194