|
Issue 004
|
|
|
AI for Operators & Founders · June 2026
AI Builder
Brief.
What happened in AI this week, and what it means for your work.
|
|
|
From the Desk
The expertise was the asset the whole time.
Jack Sykes can't code. His mother Emma spent 30 years managing food crises for some of the UK's biggest brands. Last week, Jack raised £335,000 to turn her expertise into software.
He didn't hire an engineer. He used AI tools to build a working prototype, found paying customers, and then found investors. The expertise that made the product valuable took 30 years to develop. The prototype took weeks.
This week: what Jack's story reveals about whose knowledge matters most when building with AI, what happened when Apple put Claude on 2 billion iPhones, and what a new survey says about the gap between what companies claim about AI and what they're actually doing with it.
|
Marliis Schneider
Founder & CEO, MakerSquare · Austin, TX
|
|
|
01 · The 3 Things
The AI stories worth your attention.
|
|
Founders · AI Tools
A non-technical founder turned his mother's expertise into a funded SaaS platform, without writing a line of code.
Jack Sykes had no coding background. His mother Emma had spent decades managing food crises at companies including Ella's Kitchen, Branston, and Sarson's. He watched her run incident response from spreadsheets and manual playbooks, and saw software that should exist but didn't. So he built it. Using AI-assisted development tools, he created Friday4:30, an end-to-end incident management platform for mid-market food and drink brands. Haatch led a £335,000 pre-seed round. The company already had real revenue and paying customers before the raise. Jack's explanation for how he got there: "I am not a coder, but I knew what I wanted to achieve, and there are some incredible tools that meant we could create a prototype."
|
Why it matters
The investors funded 30 years of Emma Sykes's crisis management knowledge, encoded into a product by someone who understood the problem from the inside. AI tools removed the technical dependency. The expertise was the real asset the whole time. If you have genuine depth in a field, the only thing that has changed is how much it now costs to find out whether that depth can ship as a product.
|
Read the story →
|
|
| |
|
Platforms · Infrastructure
Claude is now a default option on every iPhone. Apple just put AI in 2 billion pockets.
At WWDC on June 8, Tim Cook announced that iOS 27 will let users choose Claude, ChatGPT, or Google Gemini as their preferred AI across every Apple Intelligence feature, system-wide. Set it once in Settings and it applies everywhere: Siri queries, on-screen awareness, cross-app actions, developer tools. Apple licensed a custom Gemini model from Google for approximately $1 billion per year, but opened the door to every other frontier model at the same time. iOS 27 reaches over 2 billion active Apple devices globally.
|
Why it matters
Until this week, using Claude or any other AI model required knowing it existed and opening a separate app. Starting with iOS 27, the AI is built into the phone the same way the camera is. For anyone who has been saying "I don't know where to start with AI tools," that starting point just moved to the device already in their pocket. The barrier dropped again, and this time it dropped for 2 billion people at once.
|
Read the story →
|
|
| |
|
Research · Enterprise Adoption
86% of executives are increasing AI spending. Only 21% are actually changing how work gets done.
On June 8, Accenture and Carnegie Mellon University's Software Engineering Institute released the AI Adoption Maturity Model, built from 100-plus existing AI maturity studies, 25 executive interviews, and surveys of nearly 600 practitioners. The headline finding: 86% of C-suite leaders plan to increase AI spending in 2026. But only 21% of organizations are redesigning end-to-end processes with AI at the core. And nearly half of executives report that AI has so far delivered little impact on profit. The gap between investment and results is widening.
|
Why it matters
Most organizations are buying AI tools without changing how work actually happens around them. That's why the results are disappointing. The companies and teams that are closing the gap aren't the ones with the biggest AI budgets — they're the ones where individual operators have built real habits around how they use AI in daily decisions. The spending is not the variable. The people are.
|
Read the story →
|
|
|
|
02 · Try This Today
Find out what your expertise could actually become.
Every field has problems that existing software solves badly, because the people who built the software didn't live the problem long enough. This prompt is for finding yours — and figuring out whether there's something worth building around it.
|
Copy this prompt
I work in [your industry] and have [X] years of experience in [your specific function or specialty].
Here are 3 problems I regularly see that existing tools don't solve well:
[describe each one in a sentence]
For each problem:
1. What specific knowledge or experience is required to solve it well? What would an outsider get wrong?
2. Could an AI tool encode that knowledge to help others in my field? (Yes / Partly / No — and why)
3. What would the simplest version of that tool look like? Describe it in one sentence.
At the end: rank these by which one I should explore first, and tell me what I would need to validate before building anything.
|
A clear view of which parts of your expertise have real product potential, and a first-pass spec for the one most worth exploring. Takes under 15 minutes. The output is yours to keep, whether or not you ever build anything.
|
|
03 · Deep Dive
The build layer changed. The expertise requirement didn't.
For most of the software industry's history, turning domain knowledge into a product required a technical co-founder, months of engineering time, and surviving the translation of what you knew into a spec someone else could build from. That process filtered out a lot of people who understood the problem better than anyone — because they couldn't afford the build cost.
What changed is what building now costs. AI coding tools can produce a working prototype from a clear description of what a product needs to do. The cost dropped. The expertise requirement didn't. You still need to know the problem deeply enough to know whether what got built actually solves it.
|
"The people who understand a problem from the inside have always been the right ones to design the solution. The gap was that they couldn't build it themselves. That gap is smaller now."
|
This matters for operators and domain experts in any field, not just software. The question is no longer "do I know enough to build something?" It's "do I know this problem well enough that an AI tool could help me encode what I know into something useful for others?"
Where do the people around you still rely on calling you instead of finding an answer somewhere else? That's where the gap is. The distance between that gap and a working prototype is shorter now than it has ever been. The people closing it fastest are the ones who started with the problem, not the technology.
|
Myth of the Week
"Building an AI product requires a technical co-founder." The technical co-founder requirement was always about the build layer, not the product. AI tools have changed what the build layer costs and who can access it. The things that haven't changed: knowing the problem deeply, having the judgment to evaluate whether the solution works, and the willingness to iterate until it does. Those have always been the harder parts. They're still yours to bring.
|
|
|
04 · Tool Spotlight
Lovable.
The AI that turns your description into a working web app.
You describe what you want to build in plain English, the way you'd explain it to a new hire, and Lovable generates a working web application. Not a mockup. Not a wireframe. A live app with a real URL, connected to a real database if you need one.
The workflow: type what you want, review what it built, tell it what to change, repeat. No code editor. No terminal. Most users get to a first prototype in under an hour. The resulting app deploys immediately or hands off cleanly to a developer for production work.
It's one of the first tools we teach in week one at MakerSquare, because getting something live on day two changes how the rest of the program feels.
|
Best for
First prototypes of web tools, internal apps, client-facing dashboards
|
Price
Free tier
from $25/mo
|
Setup time
~5 minutes
|
|
|
05 · What We're Building
Twenty days to Cohort 1.
Fifteen operators, founders, and consultants. In Austin. Three deployed AI products each, built around their actual businesses, not classroom exercises.
Cohort 1 is July 6. If you've been reading these issues and thinking about what it would take to build something real around your own work, the curriculum explains what it looks like.
No coding background required.
|
Cohort 1
July 6–17
15 seats
|
Cohort 2
July 27–Aug 7
15 seats
|
Cohort 3
Aug 17–28
15 seats
|
Download the curriculum →
|
Know someone who should be reading this?
Forward it to a founder or operator who needs the signal without the noise.
Free · Every Tuesday · makersquare.ai
|
|
AI Builder Program · Austin, Texas
|
|