Agatha Christie built an empire by weaponizing competence.
In And Then There Were None, her 1939 masterpiece, she trapped ten people on an island and constructed every scene so that each suspect appeared equally guilty, equally plausible, equally worthy of suspicion. The genius was misdirection through abundance.
When everyone looks like the murderer, the truth hides in the crowd.
Ask AI for ten content ideas and you get ten equally plausible suspects.
The absolute worst outcome of a brainstorming session is getting ten good angles. Bad ideas are harmless because you delete them instantly. But good ideas are a massive trap. AI gives you a list of structured concepts with catchy hooks and logical outlines.
You pick one and start drafting. Halfway through the page, the whole framework collapses because you realize you are interrogating a mannequin. There is zero insight underneath the pretty formatting.
You need an island, a systematic way to lock these ideas in a room and eliminate them until only the ones with a pulse survive.
âââ ââ ââ â âââ
Hi, Iâm Mia. I write about building with AI the way it should be done: with a brain, a plan, and zero circus tricks.
New to ROBOTS ATE MY HOMEWORK? Start here. Want the systems? RobotsOS. Want a personalized AI roadmap? Take a 20-second quiz.
Christie put ten suspects on an island and killed them off until only the truth remained. Thatâs what weâre doing to your content ideas today. Minus the poison.
Why did brainstorming get harder when the AI got better?
This is not another âhow to generate better content ideas with AIâ piece. You have enough ideas and thatâs your problem.
Psychologist Barry Schwartz spent years studying this exact phenomenon. His work on the paradox of choice proved that flooding a human brain with attractive alternatives guarantees a miserable outcome. It triggers a cascading system failure:
Total gridlock: Staring at a blinking cursor while unable to pick a lane.
Creative resentment: Second-guessing whichever concept you finally select.
The lingering ghost: A haunting certainty that the idea you abandoned was the real masterpiece.
Sheena Iyengarâs famous jam study at Columbia confirmed the same thing in a grocery store. Customers shown 24 jam varieties were 10 times less likely to buy than customers shown 6. More options, less action.
If youâre stuck in a cycle of generating endless variations and using none of them, this is exactly why your AI feels useless.
Now apply that to your content calendar.
Before AI, you brainstormed three decent ideas and picked the obvious winner. Now you ask an LLM for help, and suddenly you have twelve. And because AI is exceptionally good at dressing up mediocrity, all twelve look publishable.
I call this plausible-idea fatigue.
You have too many ideas that look finished. Good enough, but nothing grabs you by the throat and demands to be written.
The expensive lawyer defending your worst suspect
Besides generating ideas (and a lot of them), AI also likes to polish them and dress them in the best outfit. At first glance, they all sound like good content ideas.
This is dangerous because the output has the texture of finished thinking but the substance of a first pass. You are basically dealing with a suspect who hired an expensive lawyer despite being guilty.
We also have to struggle with a compounding effect. Research on AI-generated text shows that models trained on their own output converge toward the mean. Each generation gets flatter and more generic.
Four things happen when every idea looks passable:
Plausible ideas are harder to kill than obviously bad ones. When an idea is terrible, your job is easy: you identify the flaw and you delete it. A plausible idea resists the chopping block because it did nothing noticeably wrong. It remains perfectly acceptable and could easily survive a quick inspection.
Writers choose on mood when criteria are missing. If you do not have a named hierarchy for what matters most, your decision defaults to whichever option makes you feel excited at that specific moment. You end up relying on vibes, and relying on vibes is not a sustainable editorial strategy.
Your editorial strategy needs structure, which means you need to stop prompting and start engineering context instead.
AI makes weak ideas sound stronger than they are. The thinnest idea in your list just received the exact same confident packaging as your strongest one. It becomes difficult to tell the difference without testing them.
Too many ideas create false momentum. Having ten viable options on your screen feels productive. You feel like you have completed the hardest part of the job, when you have only managed to skip the difficult work of making a decision.
The criteria hierarchy: when everything looks equally good, decide in this order
When all your surviving ideas look plausible, you need a strict hierarchy to decide what stays:
Specificity beats originality. A boring but highly specific idea will always survive over an interesting but vague one, because specificity gives you material to build upon.
Tension is non-negotiable and always comes first. An idea without an underlying tension will collapse halfway through the page. You need two things that look compatible but are not, or a universally held belief that feels just a little bit wrong.
Emotional pull beats hook strength. A hook gets someone to click, while emotional pull gets them to finish. If you have to choose between an idea with a killer opening line and no emotional weight, or an idea thatâs emotionally resonant with a weak hook, take the pull. Hooks can be rewritten in revision. Pull is either there or it isnât.
Hook can be taught, taste canât. If youâre weighing two finalists and one has your distinct taste in it (your weird angle, your specific framing, the thing only you would notice), take taste over a better hook.
The anatomy of a fragile content idea
If you canât name exactly whatâs wrong with an idea, you wonât be able to kill it efficiently.
I break ideas down into five distinct categories during an elimination round. Not all of them are inherently bad, but only one of them is worth spending your afternoon on.
Promising ideas are undeniably interesting but severely underdeveloped. They have a brilliant spark, but the structural build just isnât clear yet. You should put these in a drawer, because they might survive a brainstorm next month. Or, you could put them in the đżGREENHOUSEđż. More on that next week.
Plausible ideas sound perfectly decent on the surface, but they will never carry a full piece. They look finished in a list format, which makes them the most dangerous suspects in the room.
Fragile ideas fall apart the second you ask one hard question. If you wonder where the draft would sag and the whole concept crumbles, youâre looking at a fragile idea. They usually have a great hook and absolutely nothing else. These are the same hollow ideas that prompt most writers to start overusing AI research until they sound completely generic.
Bloated ideas try to do three jobs at the exact same time. You likely merged two thin angles together because neither felt strong enough on its own, and now neither concept has any room to breathe.
Surviving ideas remain compelling even after you apply pressure. They have specificity, tension, emotional pull, and a clear architectural trajectory you can vividly picture yourself writing. These are rare finds. You might get exactly one per brainstorm.
Brainstorming asks: what else could this be?
Elimination asks: what disqualifies this?
Agatha Christieâs drawing room is always packed with plausible suspects, but the drawing room doesnât really care about plausible.
If you want to stop writing drafts that collapse, you have to stop accepting plausible angles.
THE CHRISTIE IDEA ELIMINATION: The 4-step system for editorial execution
Here it is. The method named after the writer who understood that the truth emerges from removing everyone whoâs lying.
The brilliance of Christieâs island is the nursery rhyme hanging on the wall. Ten little soldiers, getting picked off one by one based on their own fatal flaws. We need to bring that exact energy to our editorial process. Force our ideas to earn their right to survive your afternoon.
đ The rest of this piece is for premium subscribers.
What follows: The complete Christie Idea Elimination system. Four prompts, the human checkpoints between each step, and the editorial hierarchy that prevents AI from hijacking your ideas. This is how PREMIUM ROBOTS systems are built, and they stack: each one trains your instincts for the next, until you stop needing someone to walk you through it. Thatâs what youâre subscribing to.