Most debates have more than two options
“Either we let AI run completely unregulated, or we ban it entirely.”
That isn’t a choice between two options.
It’s a choice between two extremes with an enormous middle ground being quietly erased. The person presenting it as a two-way decision is counting on the listener to accept the menu as offered and pick one of the two items on it. Everything between the extremes just disappears.
This is the false dilemma.
Also called false dichotomy, or black-and-white thinking. The fallacy presents only two options when many more actually exist. It’s one of the most effective rhetorical moves in politics because it forces the audience to pick a side in a war somebody else constructed, rather than considering the full range of possibilities. The argument isn’t between the two sides. The argument is about why those two are the only sides.
The AI debate is a perfect example right now.
People who express concerns about AI often get accused of wanting to shut the whole thing down. But what do the actual critics object to? Data centers consuming massive amounts of energy – which is a problem shared by cryptocurrency mining, streaming services, and cloud computing generally. That isn’t anti-AI. It’s an energy policy concern that applies broadly across tech. The same solution covers all of it.
Or maybe the concern is data consent.
Which is an issue with Facebook, Google, and Amazon too. Address it with stronger privacy laws across tech. Not with AI-specific bans.
Or AI art “stealing” creative work.
AI learns from general styles the way human artists do. Directly copying a specific artist’s work is still copyright infringement either way. And AI-generated art currently cannot be copyrighted in the United States, which means human artists retain a legal advantage that AI-generated work doesn’t have. That’s a real policy distinction that gets erased when the whole debate collapses into “love AI or hate AI.”
The false dilemma flattens all of these specific, addressable concerns into “ban it or don’t.”
That framing makes thoughtful policy impossible. Which is often the point. The people framing the debate this way want the middle ground to vanish, because the middle ground is where they’d lose. Extremes are easier to fight than specifics.
It shows up everywhere.
“You’re either with us or against us.” “If you’re not part of the solution, you’re part of the problem.” “Love it or leave it.” “Either we fund this program fully or we defund it.” Each of these eliminates the middle ground where most real-world policy actually lives. Each one pretends that a spectrum is a toggle. The pretense is the fallacy.
This connects to the syllogism test from Essay 13.5.
When an argument leaps to an extreme conclusion, there are usually a dozen more reasonable conclusions that follow from the same evidence. The leap isn’t inevitable. It’s one option among many, dressed up as the only option because the alternatives didn’t get named. Naming the alternatives is almost always enough to dissolve the dilemma.
The fix is to ask what other options exist.
What’s being left off this menu? Almost always, the space between two presented extremes contains multiple viable positions. Somebody who insists on a binary choice has usually done the pre-work of excluding the options they didn’t want you to consider. The move is to put those options back on the table and see how the argument holds up once the menu is actually complete.
Reality is complicated.
Anyone trying to reduce it to two choices is usually trying to prevent the listener from noticing the options they’d rather not be discussed.
Put the other options back on the menu.
Then the argument gets interesting.