The alignment problem

Beth Singlet writing for Aeon:

My stomach sank the moment the young man stood up. I’d observed him from afar during the coffee breaks, and I knew the word ‘Theologian’ was scrawled on the delegate badge pinned to his lapel, as if he’d been a last-minute addition the conference. He cleared his throat and asked the panel on stage how they’d solve the problem of selecting which moral codes we ought to program into artificially intelligent machines (AI). ‘For example, masturbation is against my religious beliefs,’ he said. ‘So I wonder how we’d go about choosing which of our morals are important?’

The audience of philosophers, technologists, ‘transhumanists’ and AI fans erupted into laughter. Many of them were well-acquainted with the so-called ‘alignment problem’, the knotty philosophical question of how we should bring the goals and objectives of our AI creations into harmony with human values. But the notion that religion might have something to add to the debate seemed risible to them. ‘Obviously we don’t want the AI to be a terrorist,’ a panellist later remarked. Whatever we get our AI to align with, it should be ‘nothing religious’.

Generalised and specialised AI

Barack Obama in conversation with Wired:

“My general observation is that it has been seeping into our lives in all sorts of ways, and we just don’t notice; and part of the reason is because the way we think about AI is colored by popular culture. There’s a distinction, which is probably familiar to a lot of your readers, between generalized AI and specialized AI. In science fiction, what you hear about is generalized AI, right? Computers start getting smarter than we are and eventually conclude that we’re not all that useful, and then either they’re drugging us to keep us fat and happy or we’re in the Matrix. My impression, based on talking to my top science advisers, is that we’re still a reasonably long way away from that. It’s worth thinking about because it stretches our imaginations and gets us thinking about the issues of choice and free will that actually do have some significant applications for specialized AI, which is about using algorithms and computers to figure out increasingly complex tasks. We’ve been seeing specialized AI in every aspect of our lives, from medicine and transportation to how electricity is distributed, and it promises to create a vastly more productive and efficient economy. If properly harnessed, it can generate enormous prosperity and opportunity. But it also has some downsides that we’re gonna have to figure out in terms of not eliminating jobs. It could increase inequality. It could suppress wages.”

Specialised AI applications are programmed to do one job; generalised AI, is programmed to do any.