How AI Is Making the Lazy Brain More Productive

“The best ideas are often the ones that are too small to bother writing down.”

For years, I have noticed something peculiar about my own thinking. Ideas rarely arrive as complete articles, theories or polished arguments. They usually appear as tiny seeds: a question, an analogy, a contradiction, a strange connection or one sentence that suddenly seems worth exploring.

Most of those seeds would normally disappear.

The brain is lazy in a very practical sense. Developing an idea takes effort. You have to retrieve related information, test the thought, find examples, organise it, write it and then edit it. A person may have the seed but not enough motivation to grow the tree.

This is where AI has become unexpectedly useful to me. I do not necessarily ask it to invent the idea. I give it the seed.

AI as an idea amplifier

A conventional view of AI productivity is that the machine should produce the finished product faster. Write the report. Summarise the research. Create the presentation. Improve the article.

But there is another, perhaps more important, use.

AI can reduce the activation energy required to develop a thought.

Suppose a person has only this thought: “Human laziness may actually be an advantage if AI can do the tedious part of thinking.” A conventional writer might leave it there. With an AI conversation, that sentence can become a question, then an argument, then examples, objections, refinements and finally an article.

The important intellectual input may still be the original seed. AI supplies the frictionless environment in which the seed can grow.

The lazy brain was always generating ideas

We often describe laziness as the opposite of productivity. But laziness also creates shortcuts. A lazy mind does not necessarily want to perform ten operations when it suspects that one will do.

That instinct can occasionally produce useful conceptual compression. Instead of accepting a complicated explanation, the brain asks: what is the basic minimum? What is the root? Can the whole system be reduced to a simpler mechanism?

My own thinking often works this way. I tend to start from the seed rather than the tree. I want the minimum structure from which the larger idea can grow.

Before AI, this created a bottleneck. A small idea might be interesting, but converting it into a coherent piece of writing required a disproportionate amount of effort.

AI changes that ratio.

The danger: outsourcing the mind

There is, however, a serious trap.

If I simply ask AI to write everything, I may become more productive while becoming less of a writer. The machine produces the article. The machine improves the argument. The machine finds the examples. Eventually I am merely the face attached to the output.

That is a version of the Ship of Theseus problem applied to intellectual work. If every component of my intellectual production is gradually replaced by AI, at what point is the resulting work still mine?

The answer cannot simply be “I gave the prompt.” A prompt that contains no original thought is not much of an intellectual contribution.

So I think the healthy division of labour is different:

• Human: observation, curiosity, judgement and the original seed.
• AI: retrieval, expansion, comparison, organisation and iteration.
• Human: selection, criticism and final ownership.

The machine should make thinking easier, not make thinking unnecessary.

Why the blog becomes more important, not less

At first glance, generative AI seems to make blogging obsolete. If AI can write a better article than most humans on almost any topic, why should anyone maintain a blog?

Because the value of a personal blog is not only the quality of its prose. It is a record of a person’s intellectual trail.

AI can generate ten excellent explanations of a subject. It cannot automatically know which peculiar observation caused me to connect two subjects on a particular afternoon. That history matters.

A blog can therefore become an archive of seeds.

Some seeds will prove wrong. Some will remain unfinished. Some will become useful ideas years later. Others may lead nowhere. But together they preserve something that a generic AI article does not: the evolution of a particular human mind.

AI can accelerate the lazy brain’s output

This may be one of the most interesting AI use cases of the next five years.

Millions of people have partially formed ideas that never become visible because the cost of developing them is too high. A person notices a pattern at work, a teacher discovers a useful way of explaining something, a trader sees an unusual market behaviour, or a student makes an unexpected connection.

Most of these observations vanish into private memory.

AI lowers the cost of extracting them. The person can provide a rough thought in ordinary language and then interrogate it. The AI can challenge the idea, find related concepts, identify weaknesses and help structure the result.

This creates something resembling a collective intellectual amplifier: not because AI necessarily invents the original observations, but because it makes many more human observations worth developing.

But AI should not always optimise for the best answer

There is another subtle problem. AI systems are generally very good at producing an apparently strong answer. But human thinking often benefits from preserving an imperfect framework long enough to understand it.

If I ask an AI to “improve this article,” I am implicitly asking it to optimise. It will naturally search for a better structure, clearer language and stronger argument.

But sometimes I do not want the best answer. I want to see what happens if we stay inside my original framework. I may want to explore a suboptimal model because that model contains the interesting idea.

This is similar to the distinction between finding the best move and understanding a particular line of play. A chess engine can choose the strongest move. A human student may instead want to understand why a weaker move fails.

For AI-assisted thinking, that distinction is crucial. Otherwise optimisation can quietly erase originality.

The basic minimum

My preferred approach is therefore simple: start with the basic minimum.

Give AI the smallest genuine intellectual input that captures the thought. Then let the conversation expand it.

• Start with the seed.
• Ask questions rather than demanding a finished answer.
• Let AI retrieve and connect information.
• Challenge the resulting structure.
• Keep the parts that actually reflect your thinking.
• Archive the final result.

This preserves something important. The AI becomes a kind of intellectual scaffolding, but the person remains the source of direction.

The future may belong to better seeds

As AI becomes better at writing, summarising and reasoning, polished output will become increasingly cheap. A thousand well-written articles may be generated in the time it once took to write one.

That will make the scarce resource something else: original observations, unusual questions and good seeds.

The competitive advantage may shift from “Who can write best?” to “Who notices something worth developing?”

That is why I do not see AI as a reason to stop writing. I see it as a reason to preserve more of the raw thinking that previously disappeared.

The blog then serves a different purpose. It is not competing with AI on prose. It is recording the seeds from which a particular mind grows its ideas.

AI can accelerate the lazy brain. But it cannot make the seed meaningful unless there is a mind willing to notice the seed in the first place.

“The future may not belong to the person who thinks fastest, but to the person who notices the right seed and knows what to do with it.”