A producer describes a move that sounds like sabotage. "I often will think of a bad idea and on purpose try that bad idea just to see where it leads."1 Not a good idea he's unsure about. A bad one, tried deliberately. The maker here is named: Rick Rubin. And he reports the results honestly. "Sometimes turns out not to be bad. Sometimes it's worse than I thought."1 So the bad idea is genuinely a coin flip — and that's the point. The value isn't in the idea being secretly good. It's in what trying it shows you.
"But the spark of seeing this bad idea through, you start recognizing where it can go."1 You run the bad idea, watch what happens, and the watching reveals a direction you couldn't have planned toward. Then comes the operative move. "This is a bad idea, but if this part of it got amplified and if this part of it went away, maybe it's more interesting — and that'll be the next iteration."1 Amplify one piece, drop another, and the bad idea becomes the seed of the next attempt. The whole thing is "a series of iterations and looking for clues, looking for bad ideas, strange ideas, dangerous ideas, and seeing what happens."1 And the foundation under all of it: "We never know until we try things what works and what doesn't."1
This is a method for generating direction when you don't have one. Normally you try ideas you believe in and avoid ones you don't. Try-the-bad-idea inverts that: you deliberately run an idea you expect to fail, because the trial produces information the avoidance never would. A bad idea you never try teaches you nothing. A bad idea you actually execute shows you — in concrete, sensory, undeniable form — exactly why it's bad, and that "why" points at what would be better.
Three things make it more than just "experiment." First, intentionality — you pick the bad idea on purpose, not by accident, which means you can do it whenever you're stuck instead of waiting for a mistake. Second, the iteration loop — the bad idea isn't the destination; it's input to a refining cycle of amplify-this, drop-that, try again. Third, the reframe of failure — when the bad idea turns out worse than expected, that's not a wasted attempt, it's a successful measurement. You now know one direction is dead, and you know it for real, not in theory.
The deepest claim is the foundation sentence: we never know until we try. This is an epistemological stance about creative work. Knowing-in-advance is treated as an illusion — and worse, a contaminated one, because any sense that you already know "is only going to be massively informed by these internal biases and well-worn patterns in your mind."1 The bad idea is a way to act before the biased foreknowledge can rule out the territory.
Why try something you expect to fail? Because the only way to know what an idea actually does is to run it, and the bad idea is the cheapest probe into territory your judgment has pre-marked as worthless. Your judgment is exactly the thing the source distrusts — "massively informed by internal biases and well-worn patterns."1 So the regions your taste rules out without testing are precisely the regions most shaped by your blind spots. The bad idea is a flashlight pointed there.
Run the probe and one of two things happens, both useful. It "turns out not to be bad" — your judgment was wrong, and you've found something you'd have skipped forever. Or it's "worse than I thought" — and now you know, concretely, where a wall is. Either way the trial converts a guess into a fact. The producer's "you start recognizing where it can go" is this conversion happening in real time: executing the idea makes its trajectory visible in a way that contemplating it never could.
Then the loop. "If this part got amplified and this part went away."1 The bad idea is decomposed — it's never uniformly bad; it has parts. Some part of it had a flicker; some part dragged it down. You amplify the flicker, drop the drag, and run it again. The next iteration is better not because you got smarter in the abstract but because the last trial gave you specific clues. The series of iterations is a search, and the bad ideas are how you sample regions of the space that pure planning would never visit. You're not looking for the bad idea to be good. You're mining it for direction.
This also explains why the move requires seeing it through — "the spark of seeing this bad idea through."1 A bad idea abandoned at the thought stage gives nothing; the clue lives in the execution. You have to actually make the worse version to read where it points. Trial is the instrument; the idea is just what you feed it.
Try-the-bad-idea is the action that The Vacant Mind Asks for Inspiration needs but can't supply on its own. The vacant mind protects inspiration when it arrives. But when nothing arrives — when the channel is dry — you can't just sit vacant forever. The bad idea is what you do in the drought: act anyway, run a probe, generate clues. The two are complementary halves of the same Rubin source: protect the inspiration you get, and manufacture direction when you get none.
It is also the resolution to a tension this source leaves open between non-interference and iteration. The vacant-mind painter says don't interfere. The bad-idea producer says interfere on purpose, repeatedly. The reconciliation is temporal: don't interfere in the first landing of an inspiration; do iterate aggressively across later versions when you're searching for one. Try-the-bad-idea is the licensed interference — deliberate tampering in service of finding direction.
It pairs with Wu Wei — Effortless Action in a way that looks contradictory and isn't. Wu Wei says don't force. The bad idea looks like forcing — but it's the opposite. Forcing is gripping one predetermined outcome. The bad idea is releasing the grip on outcome entirely: you try something with no investment in its success, purely to see what happens. That non-attachment to the result is Wu Wei wearing work clothes.
And it bypasses exactly the apparatus The Critic Blocks Creativity describes. The critic's whole function is to rule ideas out before trial. Try-the-bad-idea overrules the critic by acting before it can veto — and then uses the trial's result as better data than the critic's prediction.
The producer's account is a complete, compact methodology, and its structure rewards close reading. It opens with the foundation: "We never know until we try things what works and what doesn't."1 That's the axiom — foreknowledge is unavailable. Then the move: "I often will think of a bad idea and on purpose try that bad idea just to see where it leads."1 Note "on purpose" and "just to see where it leads" — intentional, and aimed at learning direction, not at success.
Then the honest results table: "Sometimes turns out not to be bad. Sometimes it's worse than I thought."1 Crucially, both outcomes are reported as fine. A method that only worked when the bad idea turned out good would be wishful. This one extracts value from the worse-than-expected case too — because "the spark of seeing this bad idea through, you start recognizing where it can go."1 The recognition is the yield, independent of the idea's quality.
Then the decomposition-and-iteration engine: "if this part of it got amplified and if this part of it went away, maybe it's more interesting — and that'll be the next iteration."1 The idea is treated as separable into parts with different value. The output isn't a verdict (good/bad) but a next move (amplify, drop, retry). And the framing of the whole: "a series of iterations and looking for clues, looking for bad ideas, strange ideas, dangerous ideas, and seeing what happens."1 Bad, strange, dangerous — the search deliberately includes the regions taste excludes.
The narrator then adds the epistemic kicker, generalizing the producer's practice: instead of asking "is this good enough," "accept that you don't know before you try it. And any sense that you do know already is only going to be massively informed by these internal biases and well-worn patterns in your mind."1 This is the case's spine. The bad idea is not a quirk of one producer's personality. It's an operationalization of a claim about knowledge — that creative judgment is biased toward the well-worn, and the only correction is trial in the territory the bias avoids.
You're stuck. The good ideas have run out and the work is sitting there refusing to move. Instead of staring harder, you reach for the worst idea in the room — the one your taste flinches at, the obviously-wrong move, the version you'd be embarrassed to show anyone. You don't talk yourself out of it. You build it.
Halfway through making the bad version you feel the flinch confirmed — yes, this is bad. But you keep going, because the clue isn't in the flinch, it's in the finished probe. You see it through. And now you're looking at the actual thing, not the idea of it, and something jumps out: this part is dead, but that part — that one corner — has a strange charge you didn't expect. The bad idea, fully executed, just pointed at something.
So you take it apart. Amplify the corner with the charge. Drop the dead weight. Run it again. The next version is built from the clue the bad one gave you. Maybe it's still not right — fine, it's another probe, another clue. You're not failing; you're sampling. You keep this up — bad ideas, strange ones, dangerous ones — each one a flashlight into ground your judgment had blacked out, each iteration carrying forward the part that flickered. You never knew until you tried. So you keep trying, on purpose, into the dark.
The method has broken down when:
The source is one video essay quoting Rubin and adding narrator commentary.1 [PARAPHRASED] [POPULAR SOURCE]. 🚩 SINGLE SOURCE for this specific formulation, and 🚩 MOTIVATED REASONING worth noting — Rubin is a celebrated producer with a commercial and reputational interest in his own method sounding profound; survivorship bias is live (we hear the bad ideas that led somewhere, not the thousands that just wasted studio time). The underlying logic has independent backing — variation-and-selection search, the separation of generation from evaluation, the value of cheap failed experiments — but the source cites none of it and the claim is delivered as personal wisdom.
The central tension is cost. Rubin works in a domain (music production) where trying a bad idea is comparatively cheap — a take, an hour, a rerouted plugin. In domains where a single trial is expensive or irreversible (architecture under construction, a surgical approach, a launched product), "try the bad idea on purpose" is reckless, not wise. The method assumes cheap, reversible trials and the source never says so. Open question: how do you get the information benefit of the bad-idea probe when real trials are too costly to run — does simulation, sketching, or small-scale prototyping carry the same clue-yield as full execution, or is the clue specifically lodged in seeing the real thing through?
A second tension: the method valorizes "dangerous ideas," but danger and merely-bad are different. Some dangerous ideas are dangerous because they're genuinely generative-but-risky; others are just bad with higher stakes. The source flattens these together. Knowing which dangerous ideas are worth the probe is a judgment the method claims to bypass but probably can't fully escape.
The narrator (Creative Minds, "Will") presents try-the-bad-idea as a clean illustration of his thesis — bypass the rationalizing, controlling brain; favor action over knowing.1 It fits beautifully, maybe too beautifully. He doesn't sit with the fact that this clip is in tension with the vacant-mind painter he quoted minutes earlier, who said don't interfere with the inspiration. Rubin's method is structured, repeated interference. The narrator treats both as the same anti-rationalist teaching, and at the level of "bypass the egoic mind" they do converge. But operationally one says protect-and-don't-touch and the other says probe-and-iterate-relentlessly. The convergence is on trusting action over foreknowledge; the unexamined split is on how much you should manipulate the work once it exists. The narrator never names that the two clips give opposite instructions about interference, which is the most interesting thing about putting them in the same video.
There's also a quiet convergence the narrator gets right without flagging: Rubin's "any sense that you do know already is massively informed by internal biases" is the same distrust-of-the-knowing-mind that runs through the vacant-mind clip and the empty-yourself clip. All three Rubin-adjacent moves treat the confident, knowing, planning mind as the enemy. Try-the-bad-idea is the most active expression of that distrust — not emptying the mind to receive, but acting past the mind to learn.
Business — Single-Piece Feedback Rule. The business page establishes that tight, fast feedback on small pieces of work drives learning and that the speed of the feedback loop matters more than the size of any single attempt. Try-the-bad-idea is the creative-studio instance of exactly this, and the pairing makes the hidden requirement explicit. Rubin's method only works because the loop is fast and cheap — try, see, decompose, retry, in hours. The business frame names the variable the artist leaves implicit: feedback-loop latency. The insight that emerges only by holding both: the bad-idea method is not really about courage or anti-perfectionism, as it's usually romanticized — it's about loop economics. Its value is a direct function of how cheap and fast a single trial is. Drop the cost of a trial and the bad-idea probe becomes irresistible; raise it and the same method becomes negligent. This dissolves the apparent paradox of why "try bad ideas" is wisdom in a recording studio and malpractice in surgery: it's the same method, run at radically different loop costs. The artist's mystique and the operator's loop-latency metric are describing one underlying mechanic from two sides — and the operator's frame is what tells you when the artist's advice applies.
History — Industrial Consolidation as Selection. The history page describes a population of firms subjected to selection pressure, where most variants fail and the few that survive define the surviving form — variation generated blindly, selection applied ruthlessly, direction emerging from the wreckage rather than from anyone's plan. Try-the-bad-idea runs the same variation-and-selection logic inside a single mind. The bad ideas are deliberately injected variation; seeing them through is the selection trial; amplify-this/drop-that is the inheritance of the surviving traits into the next generation. The pairing produces a genuinely useful reframe: the individual artist iterating on bad ideas is running, at small scale and fast tempo, the same blind-variation-plus-selection process that an entire industry runs slowly across decades of firm births and deaths. The history page shows the process is impersonal — no firm "knew" the right form; it emerged from selection. That impersonality is the deep justification for Rubin's distrust of foreknowledge. Just as no planner could have known which firms would survive, no artist can know in advance which idea works — both systems generate knowledge only through trial-and-selection, because the information literally does not exist before the trial. Holding both pages turns "we never know until we try" from a humble aphorism into a structural claim: in any sufficiently complex search space, foreknowledge is not merely biased, it is unavailable in principle, and variation-plus-selection is the only process that generates the missing information — whether the search runs across an industry or across an afternoon at the mixing board.
The sharpest implication. If creative judgment is "massively informed by internal biases and well-worn patterns," then your taste — the very faculty you trust most to tell good from bad — is also your most systematic blind spot. The ideas your taste rejects fastest are the ones most shaped by your conditioning, which means the regions you avoid most confidently are the regions where your originality is most likely hiding. Try-the-bad-idea is, read this way, a deliberate program of betraying your own taste to escape it. The cruel corollary: the better and more refined your taste becomes, the more confidently it forecloses, and the more disciplined you have to be about violating it on purpose. Mastery doesn't free you from the need for bad ideas. It makes them more necessary, because a stronger taste casts a larger shadow over the territory it won't enter.
Generative questions.