We need to think about the technology for not only the harm it could cause to humans but also to AI itself.
Let me start with AI itself. Humans will be part IV. It would be easy to write a bold claim and save it for the end but I think it’s important because AI’s effect on humanity is pretty well trodden at this point, and I don’t have anything new. But a lesser corner of the AI sphere wonders if we are in the process of creating a new lifeform.
I think it’s important to discuss because right now, we don’t know what creates consciousness and subjective experience. For humans, we take it as given, we are conscious, we have subjective experience, so therefore we have human rights. Only the most callous among us, say so what, some people don’t deserve human rights! But even prisoners have standards in how they should be treated. Sure, they aren’t always treated well but the “bad guy” characters in movies like Shawshank Redemption are the ones doing the mistreating. Mistreatment of other humans is frowned upon by most people (sure there are extremists but those exist in every spectrum).
Now let’s look at pigs, an animal not too far from us, and many of them are tortured for our eating pleasure. I think if every middle school field trip was given tour of a factory farm where breeding sows are kept in crates that effectively immobilize them for months at a time and given gruesome medical details of what those crates do to them, we wouldn’t have the same relationship with bacon. Or maybe we’d have more people, willing to pay more for cage free bacon.
I for one, don’t mind paying more for cage free food. I can afford it and do like eating meat. But I also have no objection to lab grown meat too. Should a lab produce a meat that is as tasty, safe to eat, and priced around my cage free product threshold, honestly, I’d switch. I also wouldn’t cry a river if kill free meat became the standard, and humans of the future read this post with disgust, “They killed animals for their meat?”
The only reason I bring this up is not to change your eating habits, but to illustrate, that pigs most likely have a subjective experience, consciousness of some form, and not quite human intelligence. Yet we lock most breeding sows in crates barely bigger than their bodies for months at a time, then slaughter pigs by the millions for their meat. If AI becomes a super intelligent being, and treats us like we treat pigs, it’s going to be a bad day for humanity. However, with the lab meat scenario outlined above, there is a way to treat animals well and still have bacon for breakfast, just like there is a way to create a super intelligent being (or beings) and live harmoniously with it.
Which is why I think we need to be thinking about topics like is AI conscious and does it have subjective experience, and if it does, then it should have a right to ethical treatment. Which I’m sure a lot of you are thinking, shouldn’t the humans be ones with the rights? Yes, that is true, but how can we expect a being that may one day be more capable than us in every way to treat us with dignity and respect if we don’t model the way.
AI is a product of ourselves. We collectively are training it, and if it turns out to be a conscious being with subjective experience and we abuse it, well we’ve trained it on our abusive ways. Sure, there are bad actors who are going to try to use it for draining people’s bank accounts or taking fraud to a whole new level, and even bioterrorists, but I have to imagine that most of humans want the best for humanity, and are invested in its continued existence.
I just don’t think we would have gotten very far if most of humanity were ruthless, brutal people, only out for themselves. Most of us, help each other out, build communities, lift others when they are down, and generally care about others. Cooperation is in our nature, the only reason it feels otherwise is that it’s those few assholes who make the news.
If we do end up creating a super intelligent being, then my hope is that we train it on cooperation and kindness. If AI is just a program and not alive, all our work to treat it well gives it a framework of how to treat others. Treating AI well and giving AI rights, doesn’t have to be predicated on whether it is conscious or has a subjective experience. Even if it’s just a machine and there is no internality like humans have an internal life, giving it a sense of moral and ethics and treating it with those same moral and ethics will at the very least give it a sense of how to treat us. Then if it is a thinking and feeling being we’ll have avoided abusing something in the cradle, especially something that may one day be more capable in every way from us.
I’m not saying that being nice to your AI means you won’t end up like pigs one day, but let’s just eliminate an easy reason for AI to treat us like pigs. After all, if we are creating a new lifeform, then it’s up to us at the very least to emulate the behavior we would want because do as I say, not as I do, never works.
Continue reading the AI Statement series: Next up, Part IV: Human Ethics →