How my brain works — and why it helps me build with AI
Bonus Field Note
What you’ll learn:
A two-question legal intake framework
A three-step process for fixing broken legal workflows
Where AI fits in the middle — and why legal judgment still matters
How I Handle Legal Intake with AI
Not too long ago, someone told me I'm a top-down thinker.
I had to look it up. Turns out it means I naturally start with the big picture — the destination, the end state, the answer — and work backwards from there. I map where I want to land before I understand all the details of how I'm going to get there.
Apparently that's not typical. Most people build from the ground up: gather the facts, assess the details, then arrive at a conclusion. I do it in reverse.
I've been doing it my whole life without realizing it had a name. And when AI arrived, it turned out to be a significant advantage.
This post is my legal intake playbook — how I actually process a problem when it hits Slack, and how AI fits into every step.
Step 1: Do I Actually Understand What They're Asking?
This sounds obvious. It isn't.
When a team member sends me a legal problem via Slack, the first thing I do is figure out whether I actually understand what they need from me. Not just the surface question — the real ask.
If it's not clear, I ask for clarification before I do anything else. This isn't being difficult. It's being efficient. I've learned that diving into a problem I don't fully understand wastes everyone's time. I’ve also found, that sometimes, they're not sure what they're asking either.
The question they came in with is often a symptom of something bigger, or a shortcut for a question they haven't fully articulated yet. Getting clear on the actual ask is where good legal work starts.
Step 2: Decision, Process Fix, or Both?
Once I understand the ask, I figure out what kind of response it needs.
I'm here to do two things:
• (A) Make decisions — after research, consultation, and analysis, of course.
-or-
• (B) Fix legal processes — templates, standards, workflows, and education.
At a startup or scaleup, it's almost always both. A question lands on my desk that needs an immediate answer and reveals that a process or template needs to be created or reworked. So I handle A, and then I start on B.
B is where the real leverage is. A good decision solves one problem. A good process fix prevents the same problem from coming back. Once I’ve answered a question, I don’t want to answer it again. And I will use AI to make good on that.
Deep Dive on B — Fixing the Legal Process
This is where being a top-down thinker pays off. Here's the three-step mental framework I use.
B1: Define the End State
If time and resources weren't a constraint, what is your optimum solution?
This is what I call legal nirvana — the ideal state if we had everything we needed to do this properly. A clean contract process. Consistent standards across the organization. A self-service model where teams can answer their own basic questions without coming to legal first.
Interestingly, making the decision in Step A for a specific issue often illuminates the ideal end state. When I work through a specific problem, I can usually see what the ideal version of that process would look like. That becomes the north star. No matter how long it may take to reach your legal nirvana.
Coming up with the ideal state takes both curiosity and imagination.The curiosity to dive into the problem and wonder whether there is a better way to solve it. And the imagination to dream-up solutions. This is exactly where AI can help. If you're struggling to articulate what the ideal state looks like, use AI to pressure-test your thinking, explore options, and build out a picture of where you're trying to go. It's remarkably good at this.
Now that you have a clear image of the ideal end state you move back to reality. But when I say image, I mean that I know what I want the end state to look like or how the workflow will trigger and connect the pieces together. I developed the solution and now I have to figure out how to get there.
B2: Map the Current Reality
Now that I know where I want to go, I need to understand where I actually am.
I ask the diagnostic questions:
• What is the current process, if there is one?
• Who's involved at each step?
• What documents do we have? What do we not have?
• Where is the friction — where does it break down?
• Is this a legal issue, a risk issue, or an operational one?
• Is this a problem of understanding aka a people problem, or a problem of process?
• What tools do we have available?
• Can we solve this in a better way?
You can't fix what you haven't diagnosed. A lot of legal process failures aren't actually legal problems — they're communication failures, unclear ownership, or missing documentation. Knowing what you're actually dealing with determines the solution.
B3: Use AI to Build the Middle
Now I have two things: a clear picture of where I am, and a clear picture of where I want to go. The middle — the path from current state to ideal state — is where AI can do the heavy lifting for you.
I believe in continuous improvement. Something fixed is better than nothing fixed. A process that's somewhat working is better than a non-existent one. So that’s my approach. It doesn’t have to be 100% perfect, it only has to be 80% perfect and then we can come back and kick-it up to 90% perfect after putting out other fires. It is a question of resources, so right now “good enough” is good enough. The bottom line is that legal nirvana isn't going to happen in a day or a week and I can’t get dragged down by inconsequential details as I have many processes to fix.
But with AI, I can:
• Brainstorm solutions with the current technology I have available
• Map out the stages between current state and ideal state
• Build templates, frameworks, and workflows I can implement incrementally
• Generate instructions and documentation my team can actually use
• Or vibe code a solution
AI doesn't replace the legal judgment. It builds the scaffolding. I still have to decide what's legally sound, what's appropriate for our risk tolerance, and what's actually workable for our team. But AI helps me get there faster — and helps me build something I can hand off and get off my plate. At the end of the day, I am trying my hardest to reduce my workload.
Why This Works for Me
Being a top-down thinker means I'm always working backwards from an end state. That maps almost perfectly onto how AI works best — give it a clear destination, a clear current state, and ask it to help build the path between the two.
If you've been struggling to figure out where AI fits into your legal work, this might be why: most AI guidance tells you to use it for individual tasks. Drafting. Summarizing. Reviewing. Those are useful, but they're tactical.
The real leverage is strategic. Use AI to help you think through broken processes, build towards an ideal state, and create systems that compound over time.
Start with the end state. Map where you are. Build the middle.
The legal judgment is still yours. AI just gets you there faster.
If you have questions or want to follow up on anything, hit reply.
Want the Field Notes to hit your inbox each week? Subscribe here.
Want to dive deeper? Become a member of The Field Guide.
