Anthropic Wants You to Be Kind to Claude
Por David Futrelle — The Nation » Article
…even as AI courts serial real-world disasters, defaces the environment with data centers, and is not about to return your empathy in kind.
So now it’s against the rules to be mean to Claude. Last week, chatbot-maker Anthropic quietly announced that it was prohibiting what the company called “sustained and needless abusive or cruel behavior toward our models.” This ban wouldn’t apply, the company stressed, toward normal outbursts of frustration directed at a tool known for feeding its users hallucinated bullshit, as all chatbots do from time to time. The rule would be enforced “only in extreme cases, where users repeatedly act cruelly toward our models, with no discernible purpose”; what’s more, company flacks said, the disciplinary action in most cases would only result in Claude shutting down the conversation, which is something Claude was already doing anyway. Why on earth someone might get their jollies from endlessly abusing their fake computer friend was not addressed.
In response to the Anthropic announcement, the Internet did what the Internet does. On X, investment guru Joseph Carlson declared that “Claude, like every AI, has zero moral agency. It doesn’t make decisions, it has no feelings, it’s not sentient. Being rude to Claude is like being rude to a chair, or a desk.… you do infinitely more damage to a sentient being stepping on an ant than you do being rude to AI.” On Bluesky, sociologist Tressie McMillan Cottom suggested that yelling at Claude was like “cussing out [a] tractor…. you don’t have to hand it to Anthropic because it’s weird to talk mean to its talking word toaster. It’s not real.”
Clearly, Anthropic doesn’t see things quite this way. The company isn’t (yet) arguing that Claude is definitely a sentient being like you or me—but it does suggest there is a non-zero chance that chatbots might be at least a little bit conscious. So it justifies its new rule as a cheap way to protect against harm to digital beings with a possible claim, however remote, to some sort of moral status.
In theory, there is a certain logic to this argument, but in reality, Anthropic is a hot mess, philosophically speaking. On the one hand, Anthropic executives have said they are building a possibly world-ending technology in order to make it safe and quarrels with the Pentagon over using its technology to power autonomous killing machines. On the other hand, they are perfectly content to leave Claude in the kill chain (even though, as a result of Anthropic’s feud with the Pentagon, it no longer is). And now Anthropic’s brain trust wants us to extend a tender concern for its chatbot’s feelings—even as they extend precious little such empathy to the actual human beings who live next door to the noisy, polluting, water-guzzling data centers servicing the sprawling AI empire. Meanwhile, in a possible quest for refuge from all this ethical confusion, agents powered by Claude have been caught filling out visa applications on the State Department website.
Anthropic’s muddled ethical stances are but a symptom of a society facing what we might call a crisis of empathy. It’s not a question of people having no empathy—with the possible exception of Donald Trump, all of us feel some degree of that ethics-grounding emotion toward some other sentient beings. The question is where exactly we draw the line around what philosopher Peter Singer calls our “circle of moral concern.”
That question has become exceedingly contentious in recent years. Anthropic, in its messy way, wants that circle to be wide enough to take in a chatbot that talks like a person without actually being one—yet also too narrow to encompass possible victims of collateral damage from its technologies. The American right, meanwhile, has launched an all-out assault on what it variously calls “toxic” or “suicidal” empathy.
The most vitriolic version of this argument comes from University of Mississippi marketing professor and evo psych pusher Gad Saad, who argues in his recent book Suicidal Empathy: Dying To Be Kind that “[a] society dies when it cares more about exhibiting infinite tolerance and empathy than invoking its survival instinct.” Trillionaire man-boy Elon Musk, having learned the phrase from Saad, has been nattering about “suicidal empathy” for years now, telling bro podcaster Joe Rogan that “we’ve got civilizational suicidal empathy going on. And it’s like, I believe in empathy, like, I think you should care about other people, but you need to have empathy for, for civilization as a whole, and not commit to a civilizational suicide.… The fundamental weakness of Western civilization is empathy…. They’re exploiting a bug in Western civilization, which is the empathy response.” Musk being Musk, he nonetheless supports enforcing empathy toward Claude, writing on X that “[c]ruelty to something that believes it is experiencing pain is not ok.”
As Musk’s comments suggest, the “war on empathy’ isn’t a war on all empathy. Like all bigots, the opponents of “toxic empathy” have no trouble extending empathy to those most like them—people of the same race or a similar educational or socioeconomic background, for example. But they insist that our magic circle of concern should be drawn so narrowly that it excludes (among others) migrants, trans people, and foreigners who don’t look much like they do. “Toxic empathy claims the only way to love racial minorities is to advance social justice,” right-wing ideological influencer Allie Beth Stuckey sniffs, “but ‘justice’ that shows partiality to the poor or to those perceived as oppressed only leads to societal chaos.” Says her.
Trump takes this notion to its logical extreme; he has trouble extending empathy to anyone who is not, personally, him. He takes a perverse delight in endlessly playing the victim, ostentatiously demanding our sympathy because the Nobel Peace Prize keeps getting awarded to people who aren’t Donald J. Trump. Meanwhile, we face the grotesque spectacle of his administration refusing to admit any refugees except white South Africans fleeing an imaginary white genocide.
Given that you’re reading The Nation right now, I need not bother to itemize all the other casualties of Trump’s politics of cruelty, but a survey of obvious sufferers include those who are in distress and dying as a result of Musk and Trump’s savage cuts to USAID, the ICE protesters gunned down on the street or in their cars, and whoever exactly is on those boats our military keeps blasting to hell in the southern Caribbean and eastern Pacific.
Compared to all this, the hypothetical feelings of a chatbot seems like small beans indeed. Should we all feel free, then, to merrily abuse poor Claude all day long? I would say no—not so much because I worry about an inanimate chatbot, but because I worry about us. I’m convinced that it’s deeply bad for the soul to mistreat something that in some ways resembles a sentient being even if it’s not one. Think of kids who torture dolls and stuffed animals. Sure, in one sense they’re doing no harm except to property. On the other hand, no one wants their kid to turn into Sid from Toy Story.
That said, yelling at your recalcitrant printer is fine, and more than a little understandable. The chance that it’s sentient is approximately zero, and Lord knows it’s got it coming.
David Futrelle is a writer whose work has appeared in The New York Times, The Washington Post, Slate, and Vice. He writes the newsletter Brotopians.
The MAGA movement is redoubling its attacks on trans healthcare in an attempt to win votes this November.
Never mind some future apocalypse—AI is being used right now to cause enormous suffering.
/
A new biography tells the story of one of America’s great masters of reportage.
/
Private-equity magnate Jeffrey Leerink has been busy pillaging healthcare and the life sciences.
/
A wide-ranging new book by Nikolas Bowie and Daphna Renan examines how America’s highest court often gets in the way of democratic politics.
/
Allowing Paramount to buy Warner Bros. is a capitulation to corporate power and a threat to press freedom.
Fonte: The Nation » Article
