Permission to Be Bad at This

Carol Dweck, the Stanford psychologist who spent decades studying how people learn, discovered something that changed education, business, and parenting. She found that people operate from one of two mindsets about their own abilities.

People with a fixed mindset believe abilities are innate. You're either smart or you're not. You're either talented or you're not. You're either a "tech person" or you're not. When they encounter difficulty, they interpret it as evidence of a permanent limitation: "I can't do this because I'm not the kind of person who can."

People with a growth mindset believe abilities are developed through effort. Difficulty isn't evidence of limitation — it's a normal part of the learning process. When they encounter something hard, they interpret it as: "I can't do this yet."

That three-letter word — yet — changes everything.

Why "Yet" Matters for AI

Most people approach AI with a fixed mindset without realizing it. They try a tool once, get a confusing result, and conclude: "I'm not good at this." They see someone else use AI effortlessly and conclude: "They get it and I don't."

These conclusions feel permanent. "I don't get AI" sounds like "I'm short" — a statement about the way things are, not the way things could be.

But using AI is a skill. Like every skill, it's bad at first, okay after a while, and eventually natural. The person who seems to "get it" didn't get it immediately. They were bad at it first. They asked dumb questions. They got useless outputs. They typed prompts that went nowhere. And then, through repetition and adjustment, they got better.

You have permission to be bad at this. That's not a consolation prize. It's the prerequisite for being good at it.

The First Pancake Principle

Every cook knows: the first pancake is always bad. The pan isn't the right temperature. The batter spreads unevenly. It sticks, or it burns, or it's undercooked in the middle. You throw it away and move on. The second pancake is better. The third is good. The fourth is great.

Nobody judges their cooking ability by the first pancake. But people routinely judge their AI ability by their first attempt.

"I tried building a website and it looked weird." First pancake. Adjust and try again.

"I asked AI to write something and it was generic." First pancake. Give it more specific instructions.

"I didn't know what to do after I set up hosting." First pancake. Google the next step, or ask AI to tell you.

The first attempt at anything is supposed to be bad. That's not failure. That's data. It tells you what to adjust. Every bad attempt gets you closer to a good one — but only if you let yourself have the bad attempt without interpreting it as a verdict on your capabilities.

What Growth Mindset Sounds Like

Fixed mindset: "I tried and I couldn't do it."

Growth mindset: "I tried and I learned what doesn't work."

Fixed mindset: "Other people are naturals at this. I'm not."

Growth mindset: "Other people have more practice. I'm getting mine."

Fixed mindset: "This is too hard for someone like me."

Growth mindset: "This is hard right now. It'll get easier."

Fixed mindset: "I made a mistake. I'm not cut out for this."

Growth mindset: "I made a mistake. Now I know one more thing."

The growth mindset doesn't ignore difficulty. It reframes it. Difficulty isn't a wall — it's a teacher. And it's a generous teacher, because it only shows up when you're learning something new, which is the only time growth happens.

The Learning Curve Is Shorter Than You Think

Here's something that might surprise you: the learning curve for using AI tools is one of the shortest in the history of technology.

Learning to use a computer in the 1990s? Months. Learning to navigate the internet in the early 2000s? Weeks. Learning to use a smartphone? Days to weeks.

Learning to use AI tools effectively? Hours.

The interface is language. You type what you want. The AI responds. You refine. That's the entire interaction model. There are no menus to memorize, no keyboard shortcuts to learn, no configuration files to edit. It's a conversation.

If you can have a conversation — and you can, because you've been having conversations your entire life — you can use AI. The only thing separating you from fluency is a few hours of practice. Not years. Not months. Hours.

Those few hours will be clumsy. Your prompts will be vague. Your outputs will be mediocre. You'll feel like you don't know what you're doing.

That's not a problem. That's the learning curve doing its job.

Giving Yourself Permission

Here's what I'm giving you permission to do. Print this list if it helps. Read it before you sit down to work.

Permission to be bad at this for a while. Your first website will be imperfect. Your first article will be rough. Your first attempt at using AI will produce mediocre results. All of this is normal and expected.

Permission to ask dumb questions. "How do I...?" "What does this mean?" "Why isn't this working?" There are no dumb questions when you're learning something new. There are only questions you didn't ask that left you confused.

Permission to make mistakes. You'll accidentally delete something. You'll publish a post with a typo. You'll choose a domain name and wish you'd chosen differently. None of these mistakes are permanent, expensive, or important. They're the normal friction of learning.

Permission to go slowly. You don't need to build an empire this week. You need to build a website. Then an article. Then another article. Speed comes later. Right now, showing up consistently matters more than showing up quickly.

Permission to not understand everything. You don't need to know what SEO stands for to publish an article that ranks on Google. You don't need to understand DNS to connect a domain. You don't need to know what CSS is to have a beautiful website. Use the tools. Understanding comes with time, and it's optional anyway.

Permission to change your mind. The business idea you start with doesn't have to be the business idea you stick with. The website you build today can be rebuilt tomorrow. Nothing is permanent. Everything is iteration.

The Identity Update

The deepest permission you need is the permission to update your self-concept.

If you've been telling yourself "I'm not a tech person" or "I'm not an entrepreneur" or "I'm not the kind of person who does this" — you've built an identity wall that no amount of information can breach.

Here's the growth mindset version: "I'm a person who's learning to use new tools." That's it. Not "I'm a tech genius." Not "I'm a born entrepreneur." Just: "I'm learning."

"I'm learning" is the most powerful identity you can hold right now. It gives you permission to be bad, permission to be slow, permission to be confused — all without any of it reflecting on your fundamental worth or capability.

Learners are allowed to struggle. Learners are allowed to ask for help. Learners are allowed to not know things. And learners — by definition — get better over time.

Adopt the identity of a learner, and the fixed-mindset voice that says "you can't" loses its power. Because learners aren't claiming they can. They're claiming they're trying. And trying is all it takes.

The Bottom Line

You are going to be bad at this at first. Every person who's good at anything was bad at it first. That's not a bug in the learning process. It's the learning process.

The only difference between the person who's flourishing with AI and the person who's frozen on the sidelines is that the first person gave themselves permission to be terrible and kept going anyway.

You have that same permission. Right now. Today.

Be bad at it. Be clumsy with it. Be confused by it. And keep showing up.

The growth is on the other side of the awkwardness. It always has been.