Bhagpuss has a post out called “It’s Real if I Say It’s Real,” with a strong argument that while people say they desire authenticity in the face of (e.g.) AI-generated music, A) people often can’t tell the difference, and B) if you enjoyed it, what does it even matter?

It was the clearest, most positive advocacy of the wirehead future I’ve ever seen in the wild.

Now, speaking of clarity, Bhagpuss didn’t advocate for wirehead in the post. Not directly. I have no personal reason to believe Bhagpuss would agree with my characterization of his post in the first place. However. I do believe it is the natural result and consequence of accepting the two premises.

Premise one is that we have passed (and are perhaps far beyond) the point at which the average person can easily differentiate between AI-generated content and the “real thing.” Indeed, is there really anyone anywhere ready to argue the opposite? Linked in the Bhagpuss’ post was this survey showing 97% of respondents being unable to tell the difference between human-made and AI-generated music across three samples. ChatGPT 4.5 already passed the classical three-way Turing Test, being selected as the human 73% of the time. Imagine that other person the research subject was texting with, and being so resoundingly rejected as human.

Then again, perhaps the results should not be all that surprising. We are very susceptible to suggestion, subterfuge, misdirection, and marketing. Bhagpuss brought up the old-school Pepsi vs Coke challenge, but you can also look at wine tasting studies where simply being told one type was more expensive led to it being rated more highly. Hell, the simple existence of the placebo effect at all should throw cold (triple-filtered, premium Icelandic) water on the notion that we exist in some objective reality. And us not, you know, just doing the best we can while piloting wet bags of sentient meat.

So, premise #1 is that it has become increasingly difficult to tell when something was created by AI.

Premise #2 is when we no longer care that it was artificially generated. For a lot of people, we are already well past this mile marker. Indirectly, when we no longer bother trying to verify the veracity of the source. Or directly, when we know it is AI-generated and enjoy it anyway.

I am actually kind of sympathetic on this point, philosophically. I have always been a big believer that an argument stands on its own merits. To discredit an idea based on the character of the person who made it is the definition of an ad hominem fallacy. In which case, wouldn’t casting aspersions on AI be… ad machina? If a song, or story, or argument is good, does its origins really matter? Maybe, maybe not.

Way back in my college days, I studied abroad in Japan for a semester. One thing I took was a knock-off Zune filled with LimeWired songs, and it was my proverbial sandbar while feeling adrift and alone. Some memories are so intensely entangled with certain songs, that I cannot think of one without the other. One of my favorites back then was… Last Train Home. By lostprophets. Sung by Ian Watkins.

So… yeah. It’s a little difficult for me to square the circle that is separating the art from the artist.

But suppose you really don’t care. Perhaps you are immune to “cancel culture” arguments, unmoved from allegations of a politician’s hypocrisy, and would derive indistinguishable pleasure between seeing the Mona Lisa in person and a print thereof hanging on your wall. “It’s all the same in the wash.”

To which I would ask: what distance remains to simply activating your nucleus accumbens directly?

What is AI music if not computer-generated noises attempting to substitute for the physical wire in your brain? Same for AI video, AI games, AI companions. If the context and circumstances of the art have no meaning, bear no weight, then… the last middle-man to cut out is you. Wirehead: engage.

…

I acknowledge that in many respects, it is a reductive argument. “Regular music is human-generated noises attempting to substitute for the wire.” We do not exist in a Platonic universe, unmoored from biological processes. Even my own notion that human-derived art should impart greater meaning into a work is itself mental scaffolding erected to enhance the pleasure derived from experiencing it.

That said, this entire thought experiment is getting less theoretical by the day. One of the last saving graces against a wirehead future is the minor, you know, brain surgery component. But what if that was not strictly necessary? What if there was a machine capable of gauging our reactions to given stimuli, allowing it to test different combinations of outputs in the form of words, sounds, and flashing lights to remotely trigger one’s nucleus accumbens? They would need some kind of reinforcement mechanism to calculate success, and an army of volunteers against which to test. The whole thing would cost trillions!

Surely, no one would go for that…