How to make sharper decisions when AI is designed to hijack your attention

Every recommendation algorithm you touch today was built by teams of PhDs whose entire job is to keep you scrolling for three more minutes. Every AI chatbot is optimized to keep you typing "one more question." And every notification on your phone is engineered to interrupt whatever deep thought you were having, right at the moment it might have turned into something useful.

None of this is an accident. It's the business model. And it means the ability to think clearly — to sit with a hard problem, weigh real trade-offs, and make a decision without flinching — has quietly become one of the rarest and most valuable skills a person can have.

This guide breaks down the actual mental models that separate people who make good decisions consistently from people who feel busy but never get anywhere. None of this requires willpower in the way most self-help content pretends it does. It requires structure.

Why does clear thinking feel harder than it used to?

It's not in your head — the environment actually got worse. Every app on your phone is competing for the same finite resource: your attention. And unlike previous generations of media, modern AI-powered feeds don't just show you what's popular. They show you exactly what your specific brain responds to, tested against millions of other people's behavior first.

That means the "just have more discipline" advice you've heard a thousand times is aimed at the wrong target. You're not fighting a static distraction. You're fighting a system that adapts to your weaknesses faster than you can build defenses against them. The fix isn't more willpower — it's better architecture around your decisions, so willpower doesn't have to carry the whole load.

What's actually going on with sleep and cognitive performance?

Here's an uncomfortable pattern worth noticing: plenty of people will spend money on nootropics, cold plunges, and elaborate morning routines chasing a tiny edge in mental performance, while running on five or six hours of sleep a night. That's backwards. Sleep isn't one input among many — it's the foundation everything else sits on. A brain running on chronic sleep debt doesn't think slower in an obvious way; it thinks worse in ways you don't notice from the inside, because your only reference point for "normal" is also degraded.

If you're serious about thinking clearly, the boring stuff comes first: consistent sleep, then everything else. Supplements and biohacks might buy you a marginal improvement. Sleep debt erases gains you've already made.

How do you use commitment devices to protect your future decisions?

One of the most useful frameworks for thinking clearly is the idea of a precommitment or "Ulysses" arrangement — named after the scene in Homer's Odyssey where Ulysses has his crew tie him to the mast so he can hear the Sirens' song without steering the ship toward the rocks. He knew his future self, in that moment of heightened desire, would be incapable of rational decision-making, so he made the decision when his judgment was sound to prevent his future self from acting on irrational impulses.

The concept has been studied extensively in behavioral economics. Precommitment involves establishing rules or constraints in advance that limit future decisions, reducing the opportunity for the impulsive self to override the rational self when temptation appears. The underlying issue is one you've probably felt yourself: the person making a decision now and the person experiencing consequences later are psychologically different selves with potentially conflicting interests.

Practically, there are a few levers you can pull to build your own version:

1. Raise friction on bad choices. Delete the app, log out of the account, put the credit card in a drawer. Alter the friction by making the alternative choices for your future self harder, with more obstacles, or the desired choice easier.
2. Make a public or financial commitment. Tell someone your deadline, or put money on the line so backing out costs you something real.
3. Borrow high-quality judgment from your past self. When you're clear-headed and rested, write down the decision you want to make and the conditions under which you're allowed to change it. Future-you, tired and tempted, doesn't get to renegotiate the terms.

The Ulysses contract works because it removes the decision from the moment when you're least equipped to make it well. If you want to go deeper on the research, Wikipedia's entry on the Ulysses pact covers the origin and modern applications in behavioral science.

How do you calculate expected value like a founder or investor?

Most people evaluate decisions based on how they feel about the outcome, not on the actual math behind it. That's why some people seem to take wild risks — sinking a lot of money into something small, or going all-in on an idea most people would call reckless — while actually being far more rational than they appear.

The mental shift is thinking in expected value: multiply the probability of an outcome by its payoff, across every possible outcome, and compare the totals. This is exactly why venture capitalists write dozens of small checks into different startups instead of betting everything on one. Each individual bet might have a low probability of paying off, but the math on the winners more than covers the losses across the portfolio. It's not gambling — it's portfolio construction applied to decision-making.

You can apply the same logic to smaller personal and business decisions:

• A $500 course that has even a 10% chance of saving you 100 hours of trial and error has a strong expected value, even though it "probably" won't work out.
• A cold outreach campaign that converts 2% of contacts into paying clients is still wildly profitable if the customer lifetime value is high enough.
• Learning a new AI tool that takes a weekend to learn but might save you 5 hours a week forever is a bet almost always worth making.

The trap most people fall into is judging a decision purely by its outcome after the fact ("that didn't work, so it was a bad decision") instead of judging the process that produced it. A well-reasoned bet that doesn't pay off is still a good decision. A reckless bet that happens to pay off is still a bad one. Separating those two things is a core skill of clear thinkers.

When should you explore versus exploit?

This is one of the oldest problems in decision theory, formalized as the multi-armed bandit problem: imagine a row of slot machines, each with a different, unknown payout rate. You must decide which slot machine to play in order to maximize your winnings, but the casino does not post the payout rates of each machine, so you must play each enough to estimate its payout rate and gain confidence in which is the best long-term choice.

That tension between trying new things and doubling down on what already works shows up everywhere in business and life. In most interesting cases, a balanced strategy is needed: if you never explore, you learn nothing new and never expand your horizon, but if you endlessly explore without exploitation, you may be throwing away all the information you've accumulated.

Here's how that plays out practically:

Early on, stay in explore mode. Test different offers, different channels, different pricing. Don't fall in love with any single approach before you've gathered enough data to know if it's actually good relative to the alternatives.

Once you find a lever that clearly outperforms the others, shift to exploit mode. Stop testing for the sake of testing. Pour resources into the thing that's working until the returns start diminishing — and only then go back to exploring.

Revisit the balance periodically. Markets change, algorithms change, and what worked last year might quietly stop working while you're still exploiting it. Build in scheduled check-ins to re-evaluate whether your current "winning lever" is still winning.

The mistake almost everyone makes is picking one mode and sticking with it forever — either endlessly chasing the next shiny tactic (permanent explore) or riding a single strategy long after it's gone stale (permanent exploit). Clear thinkers know which mode they're in and switch deliberately, not by accident.

What tools can help you think and decide more clearly?

Frameworks are only useful if you actually apply them, and a few tools make that easier:

Notion AI is a strong option for building a personal decision journal. Set up a simple database with fields for the decision, your expected-value estimate, the outcome, and what you'd do differently — then use Notion's AI features to summarize patterns across dozens of entries. Notion's getting started guide walks through setting up your first AI-powered workspace.

Claude is genuinely useful as a thinking partner for expected-value calculations. Paste in the details of a decision you're weighing and ask it to lay out the possible outcomes, rough probabilities, and payoffs in a table — it forces you to make your assumptions explicit instead of leaving them fuzzy in your head.

Freedom and similar site-blocking apps function as literal Ulysses contracts for your attention. You set the restriction while you're clear-headed (in the morning, before the day's temptations show up), and it holds even when your future self would rather scroll.

RescueTime gives you an honest, unfiltered look at where your hours actually go, which is the first step in noticing which of your daily "explore" activities have quietly turned into pure distraction with zero payoff.

How do you put this into practice this week?

Pick one decision you've been putting off — something with real stakes, not a trivial choice. Write down the possible outcomes and rough probabilities for each, the way a venture capitalist would size up a bet. Then build one small commitment device that makes the right choice easier and the wrong one harder. That's it. That's the whole system, and it compounds faster than almost any productivity hack you'll find.

The people who seem to have it figured out aren't smarter than everyone else. They've just built better guardrails around the moments when their thinking is most likely to fail — and in a world engineered to exploit exactly those moments, that might be the only edge that still matters.