This morning I built a full business website before my coffee went cold
On a café terrace in Portugal, with a ten-year-old trying to steal my açaí. And why your clients are about to start asking y
Lisa’s text
My friend Lisa is not an AI consultant. This week her clients started asking her about AI anyway.
She runs her own service business, and she messaged me saying her clients had started coming to her with questions she did not have answers to. They wanted to know how to think about using AI in their own businesses in a way that would protect their customers’ information. The questions were about ethics, about privacy, about security, about where the data goes.
AI strategy is not part of Lisa’s work. But her clients trust her, and she is who they bring these questions to.
She is not the only one fielding this. I have been getting more of these messages from friends and from clients over the last few weeks, and I think we are going to see more of them in the coming months.
Here is how I am thinking about it.
What Anthropic just did
A piece of news from this week.
A few days ago Anthropic, the company that makes Claude, announced a new model called Mythos. Here is the plain version of what happened.
Mythos is more capable than anything Anthropic has shipped before. It can do longer, more complex work on its own, taking on multi-step projects and technical builds that used to take a full team of humans going back and forth for weeks.
One of the things it can do is find security flaws inside sophisticated software and walk straight through them. Give Mythos a complex system to look at, and it can spot a weakness, design a way in, and use it, without a human holding its hand.
A model that can do this is a model you don’t just hand to the general public without handing the same weapon to every attacker on the internet at the same moment.
Which is why Anthropic held it back.
Instead, they gave early access to eleven organizations through something they are calling Project Glasswing. The list includes Google, Apple, Microsoft, Amazon, JPMorgan, Cisco, CrowdStrike, Nvidia, Broadcom, Palo Alto Networks, and the Linux Foundation.
Most of those names are the people who defend critical infrastructure, financial systems, and the software the rest of the internet runs on. Before Mythos goes to the public, they get to use it to find and patch the holes in their own systems.
The pace is getting faster
News like this matters for your business because of the pace behind it.
AI has been changing how we work for a couple of years. That is old news.
The pace of that change is getting faster, though. Claude Code and Claude Cowork completely changed how everyday people build and code, and now Mythos is a clean marker of how much things are evolving faster.
Today I was sitting in a café on the Silver Coast of Portugal with my husband and our daughter. He is a music director and producer, and he has been revamping his website to align with where his work is going in this new age. We opened a laptop together and started building.
From zero to a live site with hosting took us around forty minutes.
He looked up at me over his latte and said, “I can’t believe this. Remember how much time it took to create that website for the production company two years ago?”
I told him this was nothing compared to what is coming in the next few months.
Two years ago a project like that took a small team several weeks. This afternoon it took one sitting at a café, while our ten-year-old tried to sweet-talk me out of the biggest half of our shared açaí bowl.
What felt cutting-edge a year ago is ordinary this week, and what feels ordinary this week is going to look slow in six months. Mythos is a preview of the next step in that shrinking.
Your clients are standing inside the same current.
Their customers are starting to ask them questions they do not have answers to. The people on their team are using AI tools no one has vetted. Their industry is shifting underneath them at the same speed yours is.
When they come to you for help, a big piece of what they are looking for is someone who can help them think through what is happening.
What I would say if we were sitting together
If I were at a coffee table with you, thinking this through together, here is what I would say.
The questions Lisa’s clients are asking her are going to become the questions every client asks every service provider, and you want to have thought about your answers before that conversation lands on your table.
Four things I would ask you:
Start with your client data. Where does it live, who touches it, which tools have access to it, and what have you told your clients about how you use AI with their information? If you cannot answer that in two sentences, your clients cannot either.
Your deliverables are next. Which parts of what you hand a client were produced with AI assistance, and have you been honest with them about it? The trust cost of having this come out later runs much higher than the awkward cost of naming it up front.
Then your team. Your people are using AI tools whether you have a policy or not. You want to know what they are using, on what kind of work, and whether they understand what data they are handing to a model.
And finally, the tools you pay for. The software on your stack is adding AI features on top of itself every month. Your CRM, your project management tool, your email, your accounting platform. Each one is a door. Find out which doors are open, and what is walking through them.
The goal is not to have a final answer to any of these. Start thinking about them before a client brings one to you.
The one they come to
Because when a client does bring one of these questions to you, the reason they are bringing it to you is the same reason Lisa’s clients brought theirs to her.
Lisa’s clients did not come to her because she is an AI expert. They came to her because they trust her. She is the one they go to when they are trying to think something through.
That is what her real product is. And yours.
Every new model that ships is going to push more of these questions toward service providers like you. You did not sign up to be an AI advisor. Your clients are going to bring you these questions anyway, because you are the person they turn to when they need to think something through.
The tools will keep moving, and the names will keep changing. Through all of it, the person they bring the questions to is the person they trust in the room.
So think about these things so you are ready for their questions.
Because they will come.
xo Khaïry
Co-founder, AI District | Industry Rockstar Venture Studios AI-Powered Retention Engine. Zero-CAC Revenue. Compounding Growth.