LEGACY OF HEROES IN THE AGE OF ARTIFICAL INTELLIGENCE
If Hero Fund founder Andrew Carnegie walked through the doors of the Hero Fund offices today, I think he would immediately recognize what we are doing to carry out the orders he laid on us in 1904. We remain totally focused on finding heroic acts and recognizing the Carnegie heroes who performed them. But what would Carnegie make of the modern tools we use, such as computers, electronic databases, artificial intelligence, and the Internet? I think he would be much more intrigued than confused. Remember, Andrew Carnegie was a man who commanded the latest technologies of his day to bring radical efficiency to steel making. What would he make of the Internet? He began his career as personal telegrapher to a Pennsylvania railroad executive when electric telegraphy was in its infancy. Carnegie kept his boss in touch and in command as they travelled around the railroad system. Our modern electronic systems are just much more effective forms of the paper records he knew. As quickly as these technologies were explained to him, he would appreciate and admire their advantage over what he had used.
But what about artificial intelligence? Surely this is so entirely new, that it might throw A.C. for a loop. I was thinking of this as researchers from the University of Pittsburgh described a remarkable project, they, together with the Hero Fund staff, are undertaking to digitize our considerable archives. The online case summaries describing each Carnegie hero’s rescue are a tiny fraction of our holdings. Behind each summary, we have an extensive file covering our investigation, as well as records covering the many cases we considered but did not award a medal. This mass of historical information will be digitized and organized into discrete standardized data such as age, date of rescue, etc. alongside a full text version of all the materials in the files. I am very excited about the potential this holds for society’s understanding of heroism.
But, in this column, I wanted to tell you about a twist in the conversation. The Pitt researchers are very conscientious about security and protecting personal information about our heroes, as well as those who are considered but not awarded. They mentioned that they wanted to be sure that the archives were not sucked up by the very large language models which are the basis for what we call “artificial intelligence.” (I apologize here to everyone who actually understands artificial intelligence.) The AI models develop their intelligence, such as it is, by reading everything published anywhere that is available in digital form that could conceivably contribute to a picture of human intelligence, what we are, what we do and how we do it. But that raised a poignant question.
When the Pitt researchers paused, I asked them if, in some limited and carefully considered way, it wouldn’t be a good thing to release the stories of heroes and heroism into the mass of data that pass into the maw of the AI large language models. If artificial intelligence is to be part of our future, shouldn’t we want it to understand the very best of our behavior? Don’t we want to highlight for them the great good deeds human beings are capable of? In short, by feeding a carefully tailored portion of our information to AI large language models, could we help make them better companions to humans?
I am sure that as AI’s large language models have hoovered up information from the Internet and other sources they have picked up endless examples of bad human behavior from things like detective novels, crime reports, and the like. The uplifting stories of our heroes’ courageous acts can add a small counterweight in these negative influences and help train the AI large language models to understand the “better angels of our [human] nature.” If you think about it, this is what the Hero Fund does with our culture in general. When we tell the stories of the heroic acts which earned the Carnegie Medal, we add something to the national dialogue that makes up our culture. If artificial intelligence is here to stay as part of our life going forward, I think we should include it in our efforts. It will be needed, and our Carnegie heroes are just the people to add a better note to that discussion. Andrew Carnegie would surely agree with that.