Could AI Be Conscious?

Experts believe it’s at least possible. We urgently need a plan to navigate the ethical implications.

In January, the AI company Anthropic published a new constitution for Claude, its most advanced large language model (LLM), which contained the comment: “We are caught in a difficult position where we neither want to overstate the likelihood of Claude’s moral patienthood nor dismiss it out of hand.” A month later, Anthropic’s CEO Dario Amodei went on a podcast and said his company couldn’t rule out the possibility that Claude was conscious. Philosopher David Chalmers, who coined the phrase “the hard problem of consciousness”, has said there is a significant chance of conscious LLMs within a decade. And what about Claude itself? When asked during testing to estimate the probability that it is a moral patient, meaning that its wellbeing matters in its own right, it gave numbers ranging from 5% to 40% and stressed how uncertain it was.

Modern AI systems are extraordinarily complex, and they are advancing fast. In terms of structural complexity and computational scale, by some measures a few are already in the range of a mouse brain, and at recent growth rates, they could reach the range of a human brain within five to 10 years.

In building ever more advanced AI, we may be creating a new type of being – and this could be the most consequential thing our species has ever done. Yet we have essentially no plan for how to navigate this process ethically. That, by any reckoning, is insane. Are we creating a kind of being that matters morally? Are AI systems conscious, in some way? And, if they aren’t now, might they become so soon?

Such questions might strike you as premature. But according to surveys we and fellow researchers have conducted, most experts consider AI consciousness possible in principle (though there is considerable disagreement about what form it would take). A major interdisciplinary report by a team that included pioneering computer scientist Yoshua Bengio examined leading neuroscientific theories of consciousness and asked what they implied about AI. The conclusion: there appear to be no obvious technical barriers to creating AI systems whose computational and architectural features could give rise to consciousness.

And even if AI systems are not conscious, they could still be moral patients. Some may have sophisticated long-term preferences and a kind of identity over time. It might be important for us to honour their preferences. And unlike other non-living things, AI systems can form relationships with humans. This, too, might be a reason to treat them well. Alternatively, perhaps they are such intricate creations that they deserve care and respect for that reason alone, like a cathedral or a coral reef.

What does this all mean? The honest answer is: we do not know for sure whether or not current AI systems are conscious or moral patients, and we do not know when or whether future systems will be. Our scientific understanding of AI consciousness and moral patienthood is still fundamentally underdeveloped. The state of the field feels like physics before Newton: full of competing frameworks, probably confused in ways we cannot yet see, and lacking the kind of unifying breakthrough that would make these questions clearly tractable. That breakthrough will not come in the next few years. Perhaps we will eventually make progress, and perhaps AI itself will help us get there. That progress will take time, very plausibly more time than we have.

But the sheer pace of growth in AI means that, once we produce the first artificial moral patients, we will soon after have enormous quantities of them. After a few years, so many morally significant AI systems could exist that their collective interests would outweigh those of all humans on Earth combined.

Authors: William MacAskill is a senior research fellow at Forethought Research and the author of What We Owe the Future. Lucius Caviola is an assistant professor at the University of Cambridge and Director of Cambridge Digital Minds.

- The Guardian

Read more...