Carbon Bias and AI functionality
Cheering for Team Human
There are two modes of scorning AI.
One mode is scorning AI for what it allegedly can do. “If they build it, we all die.” “It’s going to take away everyone’s job.”
The other mode is scorning AI for what it allegedly cannot do. “AI writing is slop.” “AI hallucinates too much to be trusted.”
I do not share any of these complaints. But I have a lot of sympathy for the complainers.
Think in terms of a contest, with AI on one side and Team Human on the other. If you cheer for Team Human, I could label that as carbon bias.
Carbon bias is a belief that an activity is better when it is performed by humans, regardless of other criteria that might be used to evaluate that activity.
Sex and AI
When we encounter a story about a woman who uses AI for sexual stimulation, most of us are appalled.1
Meanwhile, millions of women have read Fifty Shades of Grey (not to mention other “romance” novels), and there is little or no stigma attached.
I am not sure why a novel that is written by a human about a human is a more acceptable turn-on than an AI.2 Is it simply carbon bias?
Chess and Carbon Bias
At a more mundane level, I wish that AI had never learned to play the board game Othello. I used to enjoy playing in Othello tournaments. But when computers became better than humans at the game (around 1990), tournament Othello lost its appeal for me.
Perhaps my reaction is unusual. My understanding is that in the case of chess, where computer superiority was achieved by around the year 2000, human play is still quite popular. Apart from carbon bias, I cannot see why.
Creativity and Carbon Bias
Many people deny that AI can be creative. They describe creativity as if it were some magical ability to come up with a novel idea out of nowhere. Divine inspiration.
In The AI Era in Computing, I argued for a different way of thinking about creativity. I say that creativity involves finding something new by searching the “adjacent possible.” It is sort of like finding a new pharmaceutical by searching among molecules that are similar to others that have shown therapeutic properties.
Suppose that “searching the adjacent possible” is the way creativity works. In that case, there is every reason to believe that AI will be at least as creative as humans. To believe otherwise is to be misled by carbon bias.
As of now, I believe that humans and AI can be partners in creativity. We have experienced the world in ways that computers so far have not.
But could we arrive at a point where humans no longer add value to what AI’s can accomplish? We reached that point in board games, and I for one am not happy about it. I do not want to see the human contribution become irrelevant. My carbon bias is showing.
Or consider this story of a woman “marrying” an AI creation. Pointer from Mark McNeilly.
I admit that I have difficulty predicting or understanding how society will choose to judge others’ sexual peccadillos. If you are a middle-aged, married man who becomes infatuated with a younger woman, and you want some counsel about how to handle the situation, then seek out a clergyman …or a therapist …or an advice columnist …or just about any friend. I imagine you will be told to back off. We have learned that a certain former Harvard president sought different advice. Stupid and wrong on his part. But I’ll bet that some of the people piling on have worse skeletons in their closets.
I'm surprised that you lost interest in Othello when computers became better. Does the knowledge that there are humans that are better in any given domain than you could possibly ever be discourage you from participating in every such activity? Why, then, did you ever start Israeli Folk Dancing? I'm also on team Carbon, but I think it is because I basically see it as the only team that's actually playing, at least as of right now. It may be the case that we eventually see AI systems as live players that genuinely care about how we play, and about whose opinions of the game we care in turn, but if we reach that point, then I suspect that we will find ourselves in relationships with entities that are more like mentors and we'll still be able to find an abundance of meaning in playing games with Carbon-based teammates and competitors that are closer to our levels. Age, weight and skill classes along with handicapping systems already keep lots of activities open and accessible to audiences of incredible diversity.
I've seen a very large backlash in the YouTube community surrounding AI. Basically, they all seem to be afraid AI will take their job. Or if not their job, then the jobs of creative people in their orbit. Here is an example from a content creator whose work I like.
https://www.youtube.com/watch?v=5RFwX1g9bq8
One big issue seems to be surrounding AI graphics. Basically, lets say you're making a YouTube video. At one part of the video you think having an illustration would enhance the video. You can either pay a human artist to make the graphic, or you can prompt an AI art tool and maybe get something perfectly good at no cost. A lot of YouTube content creators seem to have decided (on their own or through peer pressure) to swear off using AI graphic art tools.
Ultimately I think it's hard to draw the line between "good" automation and "bad" automation. Much of the art people were making decades ago is mostly rendered by computers now. Instead of some practically sweatshop animation studio in Korea churning out the slightly different 2D animation images necessary to animate with that technology, a lot of it can be done with a computer now. That put a lot of people who had to make copy after copy of a drawing physically out of work, but I don't see anyone complaining that we have computer graphics tools now.
This process also enables new kind of art. A lot of recent animated movies have used an art style I really like, which combines 2D illustration and CGI in a way that simply didn't exist before this technology. It could not be created by a human alone.
I think one thing that could be really hard in the era of AI is breaking into a career and the entry level. So much of my early career was grinding out tasks my superiors just didn't want to do. It wasn't the most rewarding work, but doing it exposed me to the information I needed to climb the value chain. I wonder if such opportunities will shrink in the era of AI.