This week I'm at the Lawn and Landscape Technology Conference in Arizona. On my way out of the San Jose airport, I had a 2-hour delay. Usually, I would be flying the team in our K&D plane, but I sold the Tiger, and the Cirrus hasn’t arrived yet, so we are flying commercial… I don’t recommend it. Anyway, while I was pacing around the airport, I stopped to watch the entire cycle of the large TV advertising slides. A lot of AI companies, actually, every single ad was an AI company. That tells me a lot about the state of the industry right now, and it's probably why you are getting sick of hearing about it. But no matter how much you try to avoid it, the Super Tsunami is hitting the shores of the trades industry. 

Every conversation I have right now eventually turns to AI. And there are two paths these conversations take. Path 1: The person shares something they are incredibly proud and excited about that they built with AI. Path 2: The person is upset, frustrated, or just all-around negative about the potential and capabilities.

I think the biggest difference between these two perspectives is the total hours spent using AI per week. People who are genuinely excited spend more time with AI; those who are negative usually dip a toe in and never actually get past the first step of programming or training their AI tool. 

So this month, I am not just going to write about AI, I am going to challenge you on it. There are two challenges, and which one is yours depends on where you stand today.

Level One: Just Start

If you are not using AI every single day, this is your challenge. Dive in. Start with your personal life, not your business. Use it to write better emails. Use it to analyze your data. Use it to brainstorm.

Here are a few ways I personally use it on a day-to-day basis: a personal trainer who builds my workouts and pushes me in the gym. A cooking consultant who plans meals, writes recipes, and builds my shopping lists. A travel agent who plans my trips smarter and faster than Drew ever could (just kidding Drew).. A communication tool that sweeps my email, Slack, texts, and DMs to tell me who I need to get back to and how I am managing my relationships. And a trend identifier that goes out every week to track what is moving in our industry.

None of this requires a technical background. You can do all of it with Claude, Claude Cowork, ChatGPT, or whatever model you prefer. What it requires is consistency. It takes time to build the habit and to train the tool. Start now and do not stop.

Level Two: Raise the Bar

If you have been doing this for the last six to twelve months or longer and you are actively deploying AI inside your business, building applications, creating automated workflows, and sending agents out to do work for you, then your challenge is different.

Audit it. Honestly, if a third party walked in today and scored the efficacy, output quality, and return-on-time of every AI workflow you have deployed, my guess is the results would humble you. I only know that because I am literally in that position today. And this doesn't mean stop innovating. It means the game has changed from deploying more to deploying better.

It is better to build one incredible agent that revolutionizes how you source new projects than to scatter seven to ten agents across the company, each doing something different and none of it well.

Tobi Lütke is the founder and CEO of Shopify. He dropped out of school after tenth grade in Germany, learned to build software as an apprentice, and turned an online snowboard shop into a company worth more than $200 billion. He recently sat down with David Senra, and it is one of the best conversations on company building I have heard in years.

Two ideas stood out to me.

The first is how Tobi educates himself. He reads constantly and calls books the closest thing to cheat codes for real life, because you can absorb the learnings of an entire career in a few hours. He changes genre every few books on purpose to widen his range. He treats learning as a discipline, not a hobby.

The second is what he did when the world changed. When COVID hit, Tobi recognized that a core assumption underneath every plan at Shopify had been invalidated. So he did not adjust around the edges. He personally reviewed every single project in the company, working sixteen-hour days, and canceled roughly sixty percent of them. He also replaced his entire executive team. He calls this pruning the decision tree. When an assumption changes, you go back to first principles and rederive everything that was built on top of it.

He also said something that has stuck with me since. One of the saddest days of his life was opening old code and being impressed by it. Because if your past work does not embarrass you, you have stopped growing.

Here is the translation for us. AI is the changed assumption of this decade. Level one is Tobi’s reading habit applied to your daily reps with these tools. Level two is Tobi’s audit applied to your AI deployments. Review everything. Kill what is weak. Go deep on the few things that actually move the business.

Most leaders respond to new technology by adding more of it. Disruptors prune, rederive, and concentrate.

You are making second quarter decisions right now. Hiring, capital, routing, pricing, tech adoption. Those decisions will either build on what Q1 taught you or ignore it.

The pace of change is not slowing. AI, autonomy, labor, capital. Going into May, the leaders who compound from here are the ones who use the next two weeks to refocus and realign their teams, not the ones who promise to circle back in November.

If the last ninety days were painful, that pain is only valuable if you convert it into a pattern. Otherwise you just paid tuition and skipped the class.

Tobi told Senra that we will look back on 2026 as the year every single business was up for grabs. I believe him. The models are improving faster than any technology I have seen in my career, and the gap between companies that use them with discipline and those that dabble is widening every quarter. What once took 6 years for a technology company to build can be done in 3 months. This isn't hyperbole; this is orders of magnitude faster than anything before. Waiting for this to settle down is not a strategy. It is a decision to fall behind.

The AI Toolbox: Five Free Giveaways

I have been building and sharing AI tools online. This month, I consolidated five of them for you. Pick one and install it this week.

1. The CEO Coach. A persona trained on the coaching methods of Bill Campbell, the Trillion Dollar Coach behind Steve Jobs, Larry Page, and Sergey Brin. Load it into Claude or ChatGPT and get coached the way the best in the world were coached.

2. The Forty CEO Questions on AI. The questions every CEO is asking about AI right now, answered by the people building it, along with a one-hour podcast walking through the data and findings.

3. Claude Cowork for Mobile. Cowork is now on your phone, which means you can delegate real work from the truck, the airport, or the job site. This guide gives you ten ways to start using it immediately.

4. The CFO Skill for Claude. Install it, and Claude starts thinking like a CFO. Restructure your P&L, pressure test your budget, and understand your numbers at a level most operators never reach.

5. The Landscape Intelligence Skill. This one is built for our industry. Benchmarking, production rates, P&L structure, estimating logic, and prompts that return answers aligned with real industry data. If you are in landscaping, this is a must-install.

We run K&D on Aspire. It is our work order system, our project costing system, our system of record, and much more. 

Here is why that matters in the context of an AI issue. Everyone wants to talk about agents and automation. Almost nobody wants to talk about the data underneath them. With clean data in a true system of record, AI stops guessing and starts knowing. I believe the difference is not an order-of-magnitude improvement. It is literally the difference between AI working and AI not working.

Aspire is building for that future, and it is a big reason we are aligned with them. Before you deploy your next agent, ask a harder question first. Does it have a system of record to stand on?

If you are interested in learning more about how we use Aspire, just reply to this email. We would be happy to help however possible!

One idea this month connects straight to the level two challenge. Pick one AI workflow or agent you have deployed and put it through an honest audit. Copy and paste this prompt into Claude or ChatGPT:

Act as an independent auditor. Here is an AI workflow I have deployed in my business: [describe it]. Here are examples of its recent output: [paste examples]. Score its accuracy, usefulness, and return on my time from one to ten. Then tell me whether I should improve it, consolidate it, or kill it, and give me the single change that would double its value.

Run it on one workflow this week. Then run it on the rest.

One line takeaway: Ten mediocre agents are noise. One great agent is leverage.

If you have windshield time this week, queue up two episodes. Jerry Schill and Steve Wheatcroft are two CEOs who built nine-figure companies with private equity backing, using completely different strategies and achieving incredible results. Listen to both and notice how differently the same mountain can be climbed.

On July 27, we are hosting the Disruptors AI Roundtable. If you are trying to deploy AI faster, the best way I know is to get in the room with other landscape and trade leaders who are doing it right now. Share what is working. Talk about which vendors you like and which ones you do not. It is an open forum. Disruptors only.

And if this issue was helpful, reply and let me know which level you are on. I read every response.