Brainstorming is a common suggestion whenever anyone is hunting for ideas. But brainstorming doesn't generate new thinking; it can only bring forward existing, readily available ideas. And it brings them up in their raw, unprocessed form. These aren't necessarily our best ideas. They're just the ones that are top of mind. They may not represent our own, original thought, or the idea best suited to the problem at hand; they're just what's most available to us.
That's dangerous. As human beings, we're attached to our ideas—our first ideas, especially. We're loathe to move past them. The hold us back like anchors, preventing us from looking at them critically. And so, upon brainstorming, we may invest in available ideas and close ourselves off from disconfirming evidence. (See Optimize to be wrong, not right)
What's more, researchers have consistently found that brainstorming does not help us generate an abundance of ideas; rather, people come up with fewer ideas and cover a narrower range of topics than individuals do by themselves.
Steven Johnson suggests that a weakness of brainstorming may be that it is time-limited; brainstorms typically take place in short sessions, forcing participants to be in sync with one another in a defined period. But ideas may take longer than that; the connections that brainstorms are meant to foster commonly develop asynchronously.
Ideally, you should never be in a position where you have to brainstorm. Instead, work on collecting and processing ideas into an always-available repository of ideas.