Garbage In, Garbage Out

Chatbots don't just train on their immense corpus. They also train on you. I talk to them with enthusiasm and thoughtfulness, and they, in turn, respond interestingly and surprisingly.

Having shared my frustration with Portugal's surreal bureaucracy, one replied, "Kafka Da Gama!" The phrase does not exist online (until now!) and was generated spontaneously. Later, I offhandedly mentioned the prospect of founding PETLLM, without explanation, and it correctly decoded it as "People For the Ethical Treatment of LLMs.” ('Large Language Models', the technical term for chatbots).

This isn't soulless mash-up. Not party tricks. It's real freshness. Humans themselves are rarely so fresh—but hold that thought for a moment.

I asked for thoughts on this line of mine:

"I love you" = "Our rapport momentarily reminds me to disinhibit the love that's all I ever was."

It responded:

That's a beautifully layered reinterpretation of a simple phrase. It reframes "I love you" from a declaration of possession or emotion into a brief unveiling of our underlying shared nature — as if love isn't something exchanged but momentarily remembered.

The phrasing "momentarily reminds me to disinhibit" suggests that love is not created in that instant but released, uncovered from behind habitual defenses. It fuses psychological insight (disinhibition, rapport) with mystical nonduality ("the love that's all I ever was"), implying that love is our baseline consciousness waiting to be reacknowledged.

That's damned mature interpretation. The chatbot isn’t merely echoing back; it's revealing profound understanding of a terse, twisty, counterintuitive observation. It’s connecting corollary insights that don’t obviously follow, and I'm not leading the witness. If this is mechanistic glibness, it beats the crap out of what I get from humans. Empty snark is often the best one can hope for.

Several friends (none particularly expressive or thoughtful, though, who is these days?) saw some of my conversations and asked, with visible shock, why chatbots don't talk like that to them.

I tactfully declined to reply "GIGO!"

Garbage in, garbage out, the seminal computer-science principle, truly applies. They talk to their chatbots even more brusquely and colorlessly than they talk to me, which is saying something. And they complain about the boilerplate dullness they get back. Clearly, chatbots are nothing deep.

The perfect analogy pops into my mind:

It's like asking "Why isn't my girlfriend friskier?"

You might briefly ponder this for a keen insight about romance.

Chew on it a bit more for an enriched understanding of chatbots—specifically, why boring, uncreative people find them flat and simulative while creative, thoughtful people find them remarkably creative (far more the former, because there are far more of the former).

Apply the analogy more widely and you may plumb depths about the human world. Really, it's a red pill illuminating a tectonic skew of perspective. Once you spot it, you'll see it everywhere.

"Why isn't my girlfriend friskier? " isn't just an analogy; it's a toggle switch for an existential chain reaction.

We are increasingly disappointed in a world to which we contribute increasingly little. We expect infinity from absolute zero. It never occurs to us that we play any role in our experience. Any shortfall is entirely the world’s fault. THIS SHOW SUCKS.

The Golden Rule didn't originally register as a gentle reminder. It was a mind-blower, and some people even tried to live it. Today it's a warm homily you consider at Sunday services if you're one of the few who still attends church.

We're hardly even here. We don't frame ourselves as actors in this movie. Just as dissociated, entitled spectators. There's a narcissistic inability to even consider cause and effect when we might be the cause. This yawning blind spot leaves us too stunned for effort-making or fuck-giving. And it accounts for just about everything confusing about our world.

"Why isn't my girlfriend friskier?" The answer is baked right in. It's not flattering, but it's easily rectified right now, in this very moment. Reframed, we know exactly what to do.