Unlock Editor's Digest without spending a dime
FT editor Roula Khalaf selects her favourite stories on this weekly newsletter.
In his 1928 short story “The Machine Stands Still,” EM Forster imagines a destroyed Earth where people inhabit isolated, button-lined capsules that provide their every need, from cold baths to literature, and are connected by an all-seeing communications machine. Even before the pandemic, this little-known story was being touted amongst technologists as a prescient vision of the fashionable web. A quote from it opens Shannon Vallor's insightful recent book.
Vallor is a philosopher of technology who spent most of her profession at Santa Clara University in Silicon Valley, before moving to the University of Edinburgh. Her first book (2016) was set within the California sun, gently emphasizing the importance of “virtue ethics” – philosophical approaches dating back to Aristotle, Confucius and Buddha that revolve around human qualities similar to courage, moral imagination, honesty and empathy – in learning easy methods to achieve today's technological age.
expands that argument with an urgent recent call to – because the book's subtitle reads – “reclaim our humanity in an age of machine considering.” There aren’t any prizes for guessing what's modified: the discharge of OpenAI's human-aware chatbot ChatGPT. Vallor's recent book is one in every of dozens published this yr grappling with a world where Forster's literary button has seemingly change into a reality. stands out for its witty, crystal-clear portrayal of the true threat posed by AI.
The writer urges us to acknowledge that ChatGPT and similar technologies based on so-called Large Language Models (LLMs) are removed from being the handmaidens of a superhuman intelligence, but merely “giant mirrors of code built to eat our words, our decisions, our art… after which reflect them back to us.” However real they could appear, it’s mistaken to confuse them with the beginnings of a synthetic general intelligence—AGI, i.e., machine sentience—because “these mirrors know no more concerning the lived experience of considering and feeling than our bedroom mirrors know our inner aches and pains.”
AI will endanger our future, Vallor argues, but not in the best way imagined by the “Western media and policy obsession” with AGI since ChatGPT launched in November 2022. Those who worry about sentient machines oppressing humans and imposing an alien and hostile value system on future generations are missing the purpose: AI is made in our image and reflects the dominant value system of today. What we should always worry about is how AI blocks our ability to reinvent our values and “prevents us from even knowing easy methods to create a future.”
The values of our affluent post-industrial society have brought us “to today's pinnacle of scientific genius,” but in addition “to the brink of planetary destruction.” The more power we hand over to machines that may do nothing but reflect these values, the less power we now have to use our own human skills and practical wisdom (what Aristotle called phronesis) to the issues before us.
Vallor acknowledges that her mirror metaphor is analogous to that of the imitative parrot—which is understood to have the opportunity to trick people into considering they understand greater than they really do. This metaphor was utilized in the now infamous 2021 academic paper on LLMs, “On the Dangers of Stochastic Parrots,” which sparked an issue that cost two of its authors their jobs on Google’s ethics team. It also marked the start of a shift in industry and government that now not include ethics—and ethicists—of their strategies for assuaging public fears concerning the dangers of AI. Vallor herself had just finished a two-year visiting research position at Google on the time.
Public policy is currently focused on the protection of AI, and a few insiders are talking concerning the “X-risk” – the existential risk – of AGI, which poses a threat as urgent as climate change. This, Vallor writes, is a “mirror fantasy” that’s all too easily put within the service of those that wish to delay motion on the “real and immediate existential threats” of a warming planet.
The key to “humanity and planetary thriving,” Vallor says, is recognizing that we humans are those with the creative ability to redirect the course of technology. Much like Shoshana Zuboff's groundbreaking 2018 book, the solutions Vallor proposes seem feeble in comparison with the threats she so eloquently lays out. As with Zuboff, it's as much as others to work out easy methods to take up Vallor's cause.
The urgency with which Vallor presents her case isn’t matched by a radical program of motion or a daring challenge to established political and economic power. But perhaps that’s the point: with their talk of sentient machines and the extinction of humanity, it’s the X-Riskers who’re opening themselves as much as the charge of hysteria. Our task is to counter with a measured and humane response.
The AI mirror: How we are able to regain our humanity within the age of machine considering by Shannon Vallor
