I once knew an exceptionally selfish person who justified his self-serving behaviour by appealing to a theory called psychological hedonism. Psychological hedonism says that the only thing that motivates humans is feeling good. We always act to maximize our own pleasure, even at the expense of others. The psychological hedonist can admit that humans sometimes appear to act selflessly. Perhaps Alice gives generously to charity, for example. But according to psychological hedonism, even Alice’s charitable giving is selfishly motivated: Alice gives only so that she can enjoy feeling virtuous and morally superior as a result. “So,” concluded my selfish acquaintance, “By always acting to make myself happier, I’m just doing what everyone does. No one is any better than me!”

While psychological hedonism may appeal to the self-serving, it’s implausible as a theory of what motivates us. There’s no denying that giving up our own pleasures to help others can result in us feeling good. But those good feelings are not usually what we’re aiming for. 

Suppose I’m offered the choice of either giving $100 to a charity to save several lives or taking a pill that will make me falsely believe I have given and saved several lives, leaving me feeling good about my generosity and also free to spend that $100 on myself. I’d choose to give that money away, even if the alternative would result in my feeling happier overall. Feeling good as a result of charitable giving might be welcome, but it’s not usually what motivates us.

A slightly different and perhaps more plausible variety of hedonism is pragmatic hedonism: the view that the only thing of value, when it comes to an individual living a good life, is having pleasurable experiences and avoiding pain. Unlike psychological hedonism, this is not a psychological theory about why humans behave as they do, but a philosophical theory about what is ultimately good for an individual.

Is pragmatic hedonism true? It’s fairly uncontroversial that, in assessing whether an individual’s life has gone well or badly, we need to look at how enjoyable or painful an existence it was. Someone who experienced little more than extreme suffering from birth to death probably hasn’t enjoyed a good life.

A classic thought experiment from philosopher Robert Nozick suggests that it is not.

The Experience Machine

Nozick asks us to imagine an Experience Machine—a virtual reality device that, once you are plugged into it, can provide you with any experience you desire. Inside the machine, you float in a tank with electrodes attached to your head. These electrodes stimulate your brain so that it seems to you as if you are experiencing a wonderful existence. It can create the compelling illusion that you are conversing with your favourite novelist, meeting an old friend, riding a bike through beautiful countryside, or having a successful career in business. Want to know what it’s like to climb Mount Everest, or swim with dolphins? The Experience Machine can generate those experiences for you. 

The Experience Machine also wipes your memory so that, once immersed in its virtual world, you won’t remember that you’ve been disconnected from reality. You’ll believe that what you experience is real.

Suppose you are offered the opportunity to plug yourself into the Experience Machine, ensuring that you will live out an exceptionally pleasurable life—a life far more pleasurable than you might otherwise expect. Nozick asks: would you want to plug yourself into the machine?

Most of us would choose not to live our lives inside the Machine. Certainly, I wouldn’t choose such a life.

Why not? Nozick suggests an answer: Because the world experienced inside the machine is illusory. True, the pleasures I enjoyed inside the Machine would be indistinguishable from those I’d experience if I really did climb Mount Everest or achieve remarkable business success. 

But my achievements would be fake. My relationships would be fake, too. The wonderful virtual spouse and children that the Machine generated for me would be virtual. And that matters. We value genuine achievements and relationships, and not merely the convincing illusion of these things. 

It seems that what matters to us is not just the distribution of pleasure and pain that we experience in life, but other things too.  We want to lead lives that are, in a word, authentic: that are connected to reality, and that really make a difference as opposed to just appearing to do so.

Nozick concludes that the pragmatic hedonist is wrong: what matters, so far as a living good life is concerned, is more than just the balance of pleasure and pain that life contains. Other fictional scenarios also seem to bear this out. Consider the movie The Truman Show, for example.

The Truman Show

The central character is Truman, who has been immersed from birth in a huge TV set filled with actors. The set looks like a 50s-style, white-picket-fenced, small American seaside town. The actors pretend to be his friends and family. One rather resentful actress plays his adoring wife. Truman, ignorant of this elaborate deception and the fact that he is in fact the star of a reality TV show, enjoys an idyllic existence, at least until he begins to notice clues that not all is as it seems. Notice that, in effect, the TV set and the cast function as a kind of experience machine: they generate for Truman a pleasant illusion that he mistakenly takes to be reality.

So now consider a doppelganger Truman—a hypothetical twin who really does live in such a small American town and really does have such an adoring wife. Let’s suppose that, from the inside, their worlds are indistinguishable. Which life is better? 

If pragmatic hedonism were true, then, as the two lives would share the exact same balance of subjective pleasure and pain, they would be equally good. Yet the life in which Truman’s wife really does love him—rather than merely resentfully pretending to—is surely far better.

The Matrix

Of course, there are circumstances under which many of us probably would choose to immerse ourselves in the Experience Machine. If the life we might otherwise look forward to is one of misery, then perhaps a virtual life is preferable. In the movie The Matrix, the character Cypher betrays his friends to escape an unpleasant reality. Cypher chooses to be put back into the Matrix—a virtual reality we take to be real—with his memory wiped, so that he can again enjoy a more pleasurable existence:

I know this steak doesn’t exist. I know that when I put it in my mouth, the Matrix is telling my brain that it is juicy and delicious. After nine years, you know what I realize? Ignorance is bliss.

However, while a pleasant fake life might be preferable to a deeply unpleasant real life, that’s not to say that pleasure is the only factor that matters.

The Stepford Wives

The novel and movie The Stepford Wives also seems to support Nozick’s conclusion. A married couple moves to Stepford only for the wife to discover the other husbands there have replaced their wives with robot duplicates that are submissive and subservient homemakers. She learns that she, too, will soon be replaced. 

The Stepford wives are also a kind of experience machine—they are robots designed to create the “tradwife” experience for Stepford’s menfolk. The husbands choose these fake wives over the real thing because they believe they’ll enjoy a more pleasurable existence. We view the husbands as monsters, and not only because they murder their wives. The husbands are also monstrous because they prefer a fake relationship with a subservient robotic wife to the more challenging real thing.

Most of us place far greater value on authentic relationships with real human beings, messy and difficult though that can be.

Discovering that we’ve been building an online friendship with an AI system designed to mimic an actual human being would leave most of us feeling, not just that we had pointlessly wasted our time, but that we would still have wasted our time even if we had never discovered the deceit. A life lived wholly within such a deception is a life largely wasted.

Experience Machines Are Already Here—Should We Be Worried?

Nozick concludes that our responses to the Experience Machine show that pragmatic hedonism is false. It’s clear that these sci-fi scenarios seem to trigger a powerful intuitive response in many of us—that a life lived within an illusion is less good than a subjectively indistinguishable life in reality. The scenarios remind us of a principle to which it seems we’re already committed: that while feeling good matters, it’s not all that matters.

While all-consuming experience machines like Nozick’s remain some way off, more modest versions are already here. In many cases, such machines are a boon.

Take chess computers. For decades now, they have been creating the experience of playing chess with another human being. In other sectors, AI systems are deployed for medical training—simulating the responses of patients with various medical conditions so that doctors can learn to diagnose more accurately.

Experience machines also serve therapeutic purposes. They include griefbots, which serve as digital replicas of the departed. We already use videos and photographs to remind ourselves of lost loved ones, so why not use a life-like digital avatar with which we can actually converse?

Robotic caregivers are also on the way. Demographic change is resulting in generations of the elderly whose need for care and companionship is unlikely to be met by real human beings. Artificial carers can provide the support and conversation these generations need. Of course, the elderly might prefer interaction with a genuine human being, but the illusion of such genuine human interaction could be of real value to someone who would otherwise be left isolated.

Machines capable of creating the appearance that we are dealing with real human beings can be used in an honest and transparent way. Users can know that they’re not dealing with a real human. However, such machines are increasingly being used to deceive us. We’re tricked into thinking that we are dealing with a flesh and blood individual when we’re not.

Evan Ratliff, a journalist, produced a fascinating podcast called Shell Game in which he describes how he created an AI version of himself, provided it with his cloned voice and phone, and then unleashed it upon customer service representatives, therapists, scammers, and even family and friends. The results were often amusing, sometimes disturbing. 

When AI-Evan phoned a friend to talk about a soccer game, Evans’ friend started to notice that something wasn’t right, but he couldn’t put his finger on it. Later, after discovering the truth, Evan’s friend said he had found interacting with AI-Evan particularly disturbing because he became convinced that there was something very wrong with the real Evan.

As AI and voice cloning technology become more powerful and more widely available, it will make us vulnerable. For example, it will be comparatively easy for scammers to target vulnerable grandparents using AI-faked grandchildren. It might sound like your desperate grandchild on the phone asking for urgent financial help, but in truth, it’s a bot created by scammers with access to your grandchild’s public social media posts and videos.

Deceptive Personal Relationships

Perhaps the most dangerous of experience machines are the fake friend, the fake therapist, and even the fake lover.

As technology develops, the AI-generated friendships we make online could be with bots designed so that we find them more engaging, less demanding, more affirming, more empathetic—and ultimately far more addictive—than any genuine human friend. Just as AI systems already lure us into doom-scrolling addictions, so they also be able to identify what we value most in our intimate relationships, endlessly refining their bots to make them irresistible. 

As we spend more time online, we become more isolated from real humans. Bots can step in and provide the friendships we need. They can be our therapists, our life coaches, even our romantic partners. Websites promise us that “Your dream companion awaits! Create your AI boyfriend, shape his look, personality, and bring him to life in one click.”

Might we become so emotionally attached and dependent on such virtual relationships that, even if we do realise we’re dealing with a bot, we can’t stop? Subscribers to an AI boyfriend or girlfriend provider might find that the price they pay is hiked astronomically once dependency on the bot is achieved, for it has become someone they just can’t live without. 

Many children are being socialized through interaction with uncaring circuitry and programming that can create the illusion of camaraderie, warmth, even love. This gives those ultimately in control of the AI the power to shape the next generation’s preferences. Attitudes toward politics, toward religion, towards women, towards brands and products, and towards other races and nationalities will all become highly manipulable in the interests of a select few. Experience machines certainly have the potential to improve our lives, but this technology will also give those who control it immense power over us—power they can, and I guess probably will, use to manipulate and exploit. 

Stephen Law is a philosopher and author based at the University of Oxford. He researches primarily in the fields of philosophy of religion, philosophy of mind, Ludwig Wittgenstein, and essentialism. He is author of numerous popular books, including The War for Children’s Minds. Read more of his essays here.