Source record
Stop trying to figure out where AI fits into your business. Build the whole system for AI, THEN figure out where humans add value. That's the play...
I started my SEO agency a little over a year ago, almost to the day, and this year we'll do over $1 million in revenue. The thing that helped us grow so fast is we already had a pretty good idea of what AI was capable of. And so I was able to build all of our systems from the ground up without understanding.
But, of course, in the last year, we've Learned a lot more about what AI is actually capable of. So I just wanna talk about how I would structure a business, any business, really, if I were starting one today. And this isn't only relevant for new business owners.
If you're an established business owner, it's important to understand the advantage that new businesses have. Starting today, this will be your competition a year from now.
As Mark Cuban recently said, in five years, there will be two types of businesses. Those that are doing a really good job of integrating AI into their business and those that have gone out of business.
There are two primary ways existing businesses are currently trying to incorporate AI into their work clothes. And they're both exactly wrong. Lazy CEOs are just turning to their team and demanding they use AI without any guidance or structure.
The more involved leaders are looking at their existing processes and trying to figure out which ones could be automated by AI.
But that is essentially guaranteed to fail. You're taking a system that You know is going to be obsolete in two to five years. And hoping that with a little polish, you'll be able to keep up with these new AI native companies, you've got to rethink the entire system.
Here's the exercise I recommend every business, new or old, go through. What would it look like if every part of your business that can be done on a computer was handled entirely by AI? What if no human in your organization was allowed to touch a computer?
What context would the AI need to do every job in your organization? What data would they need access to? That should help you figure out how you need to structure your company knowledge base.
What tools and software would it need to interact with? That'll help you figure out which MCP's or CLI's you need to integrate, or where you need to find new tools if your existing ones don't integrate well with AI.
And here's the tricky one. What wisdom, intuition, and nuance understanding with the AI need to have. This is the part most businesses are ignoring right now because they don't think it's possible to turn these things over to AI.
But it already is possible to turn a lot of this over to AI, and it's just gonna become more possible.
This tells you which processes you need to document robust decision frameworks for. And start collecting as many examples as possible of you or another expert in your Business going through that process now.
The goal at the end of all of this is not to replace all the humans in your business. I don't plan to replace any of the humans in my business. In fact, we're still hiring every month.
The idea is, instead of assuming humans are doing all the work in your business, and then finding the places where AI can add value, you assume AI is doing all the work. You build all the systems for AI, and then you identify the places where humans can add value, and then you double down in those areas, and that becomes your differentiator.
If you think your differentiator is going to be that you're really good at incorporating AI into your processes, remember, five years from now, all your competitors are going to be really good at integrating AI into their processes. Otherwise, they will have gone out of business.
But there will still be plenty of areas where we still need a human or just prefer a human. Like, I'm sure we could replace our account managers within a few years, but people just prefer a human account manager. And if our company is built with robust AI systems delivering high quality work at scale, the account managers are free to focus on what humans do best.
Our differentiator becomes having the best account managers available. And with how much value the AI is producing, those account managers only need to increase the value. By 10 or 20%, to make the financial decision a no brainer.
AI-native process design starts by asking what context, data, tools, and human judgment AI would need to handle computer-based work.
Map each computer-based workflow to required context, data access, tool integrations, and human judgment checkpoints.
This source is connected to AI operations / system design, AI operations / decision frameworks.
go through. What would it look like if every part of your business that can be done on a computer was handled entirely by AI? What if no human in your organization was allowed to touch a computer? What context would the AI need to do every job in your organization? What data would they need access to?...