HomeJournalThinking
ThinkingAttention

Are We Forgetting How to Think Because of AI?

This is not a dystopia essay — it is a diet one. We are eating pre-chewed thoughts and wondering why our jaws are weak. This post names the three things we are actively losing (attention span, draft-one courage, ability to notice a wrong-but-confident answer), and the tiny, annoying, extremely effective habits that bring them back. Your brain is a muscle. Right now it is doing fewer reps than a Netflix binge.

Siddharth PuriMarch 22, 20268 min read
Skill Loss & Learning

Are We Forgetting How to Think Because of AI?

March 22, 2026 · 8 min read · Siddharth Puri

Thinking is a verb. Verbs get stronger when you do them and weaker when you do not. AI has made skipping the verb extremely convenient.

This is not a dystopia essay. Nobody is forcing you to outsource your thinking. You are doing it voluntarily, three hundred times a day, because it feels fine and the short-term output is good. The problem is not what AI is doing to you. The problem is what you are not doing anymore.

The three things we are quietly losing

I have watched this happen to myself and to dozens of people around me. The losses are specific and predictable.

  • The ability to sit with a problem for more than eight minutes without opening a tab
  • Writing a first draft without prompting a model for the opening
  • Noticing when an answer is wrong but sounds right — because you have stopped building the instinct for wrongness

Each of these is a skill you built for years without noticing. You do not notice it leaving either.

The deep-work deficit

I used to be able to focus on a hard engineering problem for three hours. Now I notice myself reaching for my phone at seven minutes. Twelve if it is an easy problem. My attention span has halved in two years and I can time it almost to the minute.

This is not because I got lazy. It is because I trained myself to expect an answer within seconds. When the world was slower, my brain was more patient. Now my brain is behaving like a toddler at a traffic light.

The first-draft courage problem

First drafts are the hardest part of any creative work. They require sitting with "this is bad" for long enough to get to "this is slightly less bad." AI gives you a clean, confident first draft in four seconds. The short-term wins are real. The long-term cost is that your tolerance for the "this is bad" phase collapses, and with it, your ability to produce anything truly original.

Every original thing you will ever make lives on the other side of a bad first draft that you had to push through. Skip the push, skip the originality.

The "wrong but sounds right" problem

The quiet one. When you stop doing your own thinking, you stop noticing when someone else's thinking is off. AI outputs sound confident because they are optimised to. A confidently-wrong paragraph and a confidently-right one look identical on the page if you are not bringing your own judgement.

The only defence is your own independently-built mental model of the problem. If you outsource the mental model too, you are defenceless. You will ship wrong things and believe they are right.

What to do about it

These are not self-help tips. These are the specific habits I have rebuilt to get the thinking muscle back.

  • Write first, prompt second. Always. Even if the first draft is garbage, write it
  • Explain ideas to actual humans, out loud. Your brain compresses differently when a human is listening
  • Read a long-form piece per week with your phone in another room — your phone will be fine, I promise
  • Schedule one 60-minute "no AI, no internet" block per day for the work that matters most
  • Keep a paper notebook for thinking. Not taking notes — thinking. The latency of a pen makes you actually finish sentences

The quote that sticks with me

Your brain is a muscle. Right now it is doing fewer reps than a Netflix binge watcher.

A small warning

It will take longer than you expect to get back. Thinking muscles that have atrophied for two years will not rebuild in two weeks. Plan for two or three months of slight discomfort before the stamina returns. This is the same timeline as real physical training, because it is the same kind of thing.

The good news: once you rebuild it, you notice how few people around you are doing any real thinking at all. That is terrifying for the world and extremely useful for your career.

All postsSiddharth Puri

Keep reading

View all →
AI & Future of Work

Claude 3 vs GPT-5: What Changed and Why It Matters

March 26, 2026 · 9 min

Claude 3 vs GPT-5: What Changed and Why It Matters

They both claim to be the smartest thing ever built, and both demos look suspiciously similar. This is a ground-level look at how Claude 3 and GPT-5 actually differ in reasoning depth, long-context reliability, code quality and tool use — plus a blunt cheat sheet for which one to pick for which job. Written in English, without the benchmarks theatre.

AI & Future of Work

Will AI Really Replace Developers or Just Upgrade Them?

March 18, 2026 · 8 min

Will AI Really Replace Developers or Just Upgrade Them?

The internet has been burying developers every year since 1998 and we keep showing up for breakfast. Here is the honest split — which parts of the job AI genuinely eats (boilerplate, docs, test scaffolding, Stack Overflow archaeology) and which parts quietly get harder and more valuable (product judgement, architecture, ambiguity). Short answer: it replaces the parts of your job you hated, and the parts that pay you get more fun.

AI & Future of Work

Jobs That Will Survive the AI Revolution (And Why)

March 8, 2026 · 9 min

Jobs That Will Survive the AI Revolution (And Why)

Forget the "creative vs repetitive" framing — the real line is "work customers trust a machine with vs work they do not." This post maps three tiers: jobs that will stay deeply human (health, founding sales, investigative journalism, skilled trades), jobs that will transform rather than disappear (design, engineering, teaching, support), and jobs that are quietly becoming the best bets of the decade. A calmer, less LinkedIn-flavoured take on the next ten years.