Back to the blog

Why you forget your best ideas, and what actually stops it

Wondering 18 August 2026 6 min read

The thought arrives in the shower, in the car, halfway through an argument. By the time you could write it down it has gone, and you are left with the shape of it and none of the content.

The particular cruelty is that you can still remember it was good. You know roughly what it was about. You know it connected two things you had never put together before. What you cannot do is reconstruct it, and the harder you chase it the further away it gets.

Most advice about this is about remembering harder. That advice is aimed at the wrong problem, which is why so little of it works.

It is not a memory problem

The usual fixes are memory fixes. Repeat it to yourself. Make an acronym out of it. Say it three times. Try harder to hold on.

All of that assumes the idea was stored and then lost. Usually it was never stored at all. Memory needs a moment of attention to encode something, and the moment an idea arrives is almost never a moment you have attention to spare.

Look at where good ideas actually turn up. In the shower. Driving. Walking somewhere. Falling asleep. In the middle of a conversation you are supposed to be having. These are not coincidences: they are the times your mind is loosely occupied and free to wander, which is exactly the condition that produces the connection in the first place.

They are also, without exception, times when your hands are busy and your attention is committed elsewhere. The circumstances that generate the idea are the same circumstances that prevent you from doing anything about it.

The problem is not that you forgot. It is that capturing the thought cost more than the thought was worth at the time.

Every capture has a price, and you pay it up front

Think about what writing something down actually asks of you. Dry your hands. Find the phone. Unlock it. Open an app. Decide which app. Pick a notebook or a folder. Start a note. Give it a title, or feel vaguely bad about not giving it one. Type the thing on a keyboard the size of a matchbox. Decide where it belongs.

Every step is small. The total is not, and you are asked to pay it in the two or three seconds before the idea starts to fade.

So you do the arithmetic without noticing you are doing it. The idea is probably good. Getting it down will take a minute and interrupt what you are doing. You decide you will remember it. You do not remember it.

People take this as evidence that they are disorganised. It is much more often evidence that the system is asking for more than the moment can give. A capture system that requires you to stop what you are doing will be used when you are at a desk and abandoned everywhere else, which is where the ideas are.

The test worth applying to anything you use: can you get a thought out of your head in under five seconds, without stopping, without looking at the screen, and without deciding anything? Almost nothing passes it.

What actually works

Three changes, and they are all about lowering the price rather than raising your discipline.

1. One place, not the right place

Deciding where something goes is the step people quietly abandon first. It looks like organisation, but at the moment of capture it is a tax, and it is worse than it looks: you often do not yet know what the thought is. Filing it correctly requires understanding it, and understanding takes longer than the idea will survive.

So have one place that everything goes into, regardless of what it is. A shopping list and a business idea and a worry about your father can share an inbox for an hour without anything bad happening. Sorting is a separate job, done later, by someone with more attention than you have right now.

2. Say it, do not type it

Most people speak at somewhere around 130 to 150 words a minute. Most people type on a phone at around 35 to 40. Talking is roughly three or four times faster, and that ratio alone would make it worth doing.

The bigger reason is that speaking works in the situations where ideas actually arrive. You can talk with wet hands. You can talk while driving, while walking, while cooking. You do not have to look at anything.

There is a third benefit that surprises people. When you type, you compress. You write "app idea, finance" because typing the whole thought is too much work, and three weeks later that note means nothing to you. When you speak you tend to say the whole thing, including the reason it seemed to matter, and the reason is usually the part worth keeping.

3. Capture more than you think is worth capturing

Judging whether an idea is any good takes longer than recording it. That is the whole argument. If you filter at capture time you spend more effort on the filtering than you would have spent just saving everything, and you will be wrong about which ones mattered.

You are also a bad judge in the moment. Ideas that seem thrilling at 2am are often nothing. Ideas that seem obvious and boring turn out to be the ones you keep coming back to, and you only learn that by having a record of coming back to them.

The half of the problem almost nobody solves

Here is where most advice stops, and where most systems quietly fail.

Getting things down is only half of it. A capture system that only captures becomes a graveyard: 400 voice memos named by date, a notes app with 900 untitled entries, a document you stopped opening in March. Everything is in there. None of it is reachable. You have successfully written down a great many things you will never read again.

The value was never in the individual thought. It is in the pattern. Any one note about being tired of your job is not information. The fact that you have written some version of it eleven times since February is information, and it is not something you would ever have noticed on your own, because each time it felt like the first time.

This is the part that is genuinely hard, and it is why "just write it down" is incomplete advice. Search only helps if you remember there is something to search for. Folders only help if you filed it correctly, which brings back the tax you were trying to avoid. What you actually need is for the pile to read itself back to you.

What to do this week

If you want to change one thing, change the price. Pick a single place to put everything. Make it reachable in one action from a locked phone. Speak instead of typing. Stop deciding what deserves to be kept.

Then, and this is the part to actually commit to, put fifteen minutes somewhere in your week to read back what you captured. Not to organise it. Just to read it. The first time you notice you have had the same thought four times without realising, you will understand what the system is for.

Both halves are what Corteks is

Say the thought out loud and it is saved. The AI writes the title, the summary and the filing, so there is no folder to choose and nothing to decide at the moment you have the least attention to spare. That is the first half.

The second half is the one this article says almost nobody solves. Every week Corteks reads your own thinking back to you: what you returned to, what you have been circling, the question you have asked four times without noticing. The pile reads itself.

Private beta on iOS. We let people in a handful at a time.

No launch spam. One email when it is your turn.