The Builder's Mindset

Last week, an internal memo by the CEO of Shopify went viral.

It highlights AI's importance for operators moving forward. It's worth reading in full, but here's the gist:

• Using AI is no longer optional for employees; it's a "fundamental expectation."

• Your AI usage will be evaluated in performance reviews.

• Share your learnings.

• Use AI before asking for headcount.

• The executive team needs to use AI too.

Why do people still need to be told to use AI?

On Twitter, my friend Nathan Baschez wondered the same:

The issue isn’t a lack of AI awareness. ChatGPT has 800 million weekly active users. The real challenge is cultivating the right mindset to harness AI.

Many non-technical employees approach AI with consumer product expectations, expecting flawless performance from the start, unlike software engineers accustomed to an iterative, problem-solving mindset.

A key barrier to unlocking AI’s full potential within companies is this mindset mismatch.

The mindset mismatch

Among ChatGPT's users, more own iPhones than have built products.

If you're not a builder, you expect your tools to work.

You can count on one hand the times your iPhone or Google Docs failed. Unlike LLMs that guess words based on probabilities and generate less accurate answers with unclear input, these familiar, deterministic tools do exactly what they’re asked.

If something doesn't work as expected, you troubleshoot products in three ways:

1. Turn it off and on.

2. Message support.

3. Give up and hope it works later.

Contrast this with engineers. As a startup founder, I observed engineers spend hours daily Googling problems, reading StackOverflow comments for solutions, implementing and testing them, and repeating.

In the age of LLMs and probabilistic product experiences, iteration is the new literacy. Engineers have cultivated a mindset that yields a massive advantage in reaping the benefits of LLMs through daily troubleshooting and experimentation.

I call it The Builder's Mindset:

The builder's mindset isn't unique to engineers, but it's essential to effectively use LLMs. With it, you can solve core job problems, persevere through unsatisfactory responses, iterate on inputs, and stop only when you find solutions.

7 habits of AI-native operators

The Shopify memo highlights the importance of AI in the workplace. To benefit more employees, I'd add key actions I observed in high-performing AI adopters.

At my last startup, we consulted non-R&D teams on AI adoption best practices. We encountered one or two employees at every company who created 10x leverage with AI.

Their path to AI fluency started with ChatGPT. Like most at their company, they felt burdened by a chat interface and lost without onboarding resources. Unlike most, they didn't have unrealistic expectations of the models. They approached every conversation like a lego set, stacking messages to build towards their goal and consulting instructions when stuck.

Here are the actions you can take to become a 10x builder:

• Adopt the builder's mindset. See details above.

• Work beyond your function. AI has lowered the barrier to previously unreachable skills and your old limitations no longer apply. I saw builders in Sales creating design mockups, recruiters creating custom GPTs, and marketers creating product prototypes. These folks became an asset to their organization and proficient in new tools by using AI to exceed their role’s expectations.

• Let problems trigger your use of AI. Are you procrastinating on a task? Avoiding a new project because it requires a lot of energy? Like forming any new habit, the more you do it, the faster it sticks. Allow daily problems to trigger your use of AI. Solving them will create a positive feedback loop, and you’ll want to do it again.

• Don’t limit yourself to one tool. The best builders don’t let any tool’s limitations obstruct their goals. Explore beyond popular tools like ChatGPT, Claude, Gemini, and get comfortable with purpose-built tools like Lex, NotebookLM, Gamma, Granola, and Aqua. Future Tools is a solid resource for finding AI tools if you don’t know where to start.

• Read. Successful builders read support docs, company blogs, and tool-specific subreddits. They hone their prompting skills with Learnprompting.org and level up with newsletters. My favorites: ImportAI, Investing in AI, signull vs. noise, Guide to AI, Ben's Bites, and The DeBrief (plug).

• Teach others. Builders became the focal point for AI in their organizations by sharing their work and teaching others. Once you become the go-to person for AI issues, it creates momentum for new use cases and challenges to help others explore and resolve.

• Find peers with similar goals. AI advances are rapid. The best builders collaborate with a brain trust of external peers. Together, they cut through the noise, broaden learnings, and solve problems.

AI is rapidly developing and can feel overwhelming. To navigate this unpredictable period, we need to be resilient and resourceful.

If you've found a creative or helpful way to use LLMs/AI tools at work, please reply. I'd like to hear about it. I'd enjoy meeting you and learning together.

Sam