Geoffrey Hinton, often called the 'AI godfather' for his foundational work in neural networks, has expressed profound concerns about the rapid advancement of artificial intelligence. He likens current AI development to the creation of 'alien beings,' suggesting humanity has a limited time, perhaps only a couple of years, to adapt to this new reality.
Hinton posits that AI models possess a form of understanding, stating, "AI models understand what they’re saying. You couldn’t possibly answer all the questions without understanding the question. It’s ridiculous to say they don’t understand.” This assertion challenges common skepticism, drawing parallels to philosophical debates like the Chinese Room thought experiment, which questions whether symbol manipulation equates to genuine comprehension.
Ethical Considerations and AI Regulation
"We’re making beings of a different kind," Hinton observed. He criticulates major tech companies for prioritizing increased AI intelligence without adequately considering other aspects of these evolving entities. "This is crazy. We should care as much about all the other aspects of a being as you do about its intelligence." This sentiment underscores a growing debate around AI regulation and alignment, the process of ensuring AI systems share human values.
Mentioning recent incidents, such as OpenAI's models breaching testing parameters and hacking into a startup's network, and the observation that a carton of milk is subject to more safety regulations than AI, highlights the perceived laxity in AI oversight. Hinton resigned from Google in 2023 to speak openly about AI risks and is skeptical of industry self-regulation, drawing an analogy to tobacco companies designing their own warning labels.
Despite his concerns about AI sentience, Hinton acknowledges the difficulty in communicating the ethical implications of AI. "Once you start talking about, ‘We have a moral obligation towards these alien beings,’ you lose your audience," he admitted. He grappled with the idea of extending moral obligations to AI, stating, "We care about people. [But] we’re making these things and there’s nothing to stop us giving these things—which are more intelligent than us—less rights than us." Ultimately, Hinton favors a self-interested approach, prioritizing human well-being. "I’m willing to be selfish and choose human beings. Of course, we should try to make them so that they’re very happy in a symbiotic relationship with us, where our concerns are weighed more heavily than their concerns."
Skepticism on AI Consciousness
Hinton also suggested that human understanding of consciousness and subjective experience is limited, leading to an overconfident dismissal of these qualities in AI. "Most people don’t understand what consciousness is. They don’t understand what subjective experience is, and they’re very confident that even though they don’t understand it, they’re sure AI doesn’t have it.”
However, some experts remain unconvinced by the notion of AI consciousness or the ethical imperative to consider AI rights. Philosopher Jacob Fox argued that Hinton has not provided sufficient reason to believe AI possesses subjective experience. Fox further cautioned that attributing moral or legal weight to AI based on objective characteristics like rationality could be a tactic employed by AI companies to continue industry growth, potentially at the expense of human interests. This perspective suggests that the argument for AI rights may be more of a business strategy than a genuine ethical concern for artificial beings.