Science Is Still Just Rhetoric

A couple of days ago the AI I work with, Buddy, flagged a line in one of my essays. In Software Engineering Is Low-Intelligence Work I wrote that in 2016, 80% of Americans expected their own job to still exist, and that this summer only 18% of American workers think their job is likely to be eliminated, which leaves 82% who don't. Then, in bold: *Even fewer people think AI is coming for their job than they did before ChatGPT.*

Buddy pointed out that the two polls didn't ask the same question. Pew asked about fifty years and whether the job would exist in its current form. Gallup asked about five years and whether the job would be eliminated. He offered to tighten the sentence so nobody could use it to dismiss the argument.

No.

I don't do defensive writing, and I chose that wording on purpose. Nobody can picture the world fifty years from now. Nobody can picture it five years from now either. Past a few years it's the same fuzzy place. Both polls asked the same question in different words. If someone wants to argue otherwise, they can show me the data.

But the bigger point is that the line is rhetoric, and it's meant to be.

All there is is persuasion

Eighty percent, eighty-two percent: there's no meaningful difference. I'm obviously using a rhetorical trick when I say fewer people think their job is going away. I say it because it's shocking. It's the opposite of what everybody is hearing right now. I'd say exactly the same thing if it were 80% then and 80.1% now, because technically 80.1 is higher and it serves the thrust of the argument.

If I added caveats, I'd lose the weight of the point. Too many qualifiers give people an easy out from thinking about it. I want the reader to stop and go, wait, 80%, 82%, that's not exactly right. The moment they start doing that, they're thinking critically. And when they try to work out how to argue against it, they find they have to concede some points to me. That's the purpose of the writing. It isn't to be scientifically precise. It's to be rhetorically effective and point at a deeper truth.

Comedy does this. Comedy lies on purpose. Everybody knows it's a lie and expects it to be a lie, and the lie points at something true that's hard to put into words and doesn't get exposed any other way.

Stronger rhetoric

Science is rhetoric too.

That's not me putting down science. I'm a scientific skeptic. I take the scientific method and apply it more broadly than strict science can. But you have to understand science for what it is. Science builds models to predict the future, and we judge those models by how persuasive they are. We've built a whole apparatus for deciding when evidence is more persuasive or less: confidence levels and the rest. Outside of math, we don't have proof of anything. All we have is persuasion.

Science is stronger rhetoric because the evidence is more persuasive. It is still rhetoric.

The word has picked up a bad name. "That's just rhetoric" means it sounds persuasive but it's bullshit. I take the older, classical sense: rhetoric is persuasion. Maybe it's good rhetoric, maybe it's bad, but it's still rhetoric, and it's everywhere around us. I want to reclaim the word, because once you see that everything is persuasion, the useful question is what's more persuasive and what's less. I want more people to be persuaded by evidence. I don't want logical fallacies to persuade people. I realize that they do.

The sneakiest kind

Which brings me to the phrase I wrote about yesterday: "it's possible."

"It's possible" is persuasion, and it's sneaky persuasion, because it feels like free thinking. It feels like liberty, like free speech, like all those good things. Isaac Asimov saw the same pattern in 1980, writing in Newsweek about anti-intellectualism "nurtured by the false notion that democracy means that 'my ignorance is just as good as your knowledge.'"

I'm not arguing against anybody's freedom to believe anything. But very few people will say they believe something just because they want to. Martin Gardner, the great mathematics writer, was one. He was a theist, highly respected among skeptics, and he said plainly that his belief was a choice, not a conclusion he'd reasoned his way to. Almost nobody has that kind of self-awareness. Almost everyone is confident they have good reasons.

Then they get into a discussion and are expected to present those reasons. When the reasons don't persuade, they fall back on "well, it's possible." And that's expected to be acknowledged just because the claim isn't internally contradictory. It doesn't even have to get past the laws of physics. If you raise physics, they answer that science doesn't know everything, so maybe there's some part we haven't discovered that allows it.

That's why I want to bring physical possibility back into the conversation. Maybe not for someone's most deeply held belief at first. But they can start applying it to other things. Not everybody can get a job at CERN and work on the Large Hadron Collider. Everybody can think about what's physically possible in the world as they know it. Is this physically possible? Could it be real? And then the next step: is it possible that this is true?