Aggressively Human: Online Business in the Age of AI, Algorithms & Automations

AI coding tools have made it possible for anyone to build an app by just describing what they want. And solopreneurs are doing exactly that—from quiz tools to iOS apps to full CRMs. However, there’s a lot more to this conversation.
In this episode, we’re joined by Joe Casabona of the Streamlined Solopreneur. Joe has a master’s in software engineering and spent 20+ years as a developer before pivoting to help solopreneurs build systems that let them actually take time off. He’s been deep in the vibe coding world—building iOS apps, beta reader tools, and more—and has strong opinions about what’s worth your time and what isn’t.
We get into what vibe coding actually is, which tools people are using, and how code goes from running in a sandbox to something real people can access. This isn’t a “vibe coding is the future” episode. It’s an honest look at the tradeoffs—security, cost, opportunity cost, and the sneaky way tinkering can start to parade as productivity.
* What vibe coding actually is, and which tools solopreneurs are talking about most (Lovable, Claude Code, Cursor, Codex, Replit)
* How code goes from a thing running on your computer to something actual humans can access and use
* The types of projects that are good candidates for vibe coding—and the ones you should probably leave alone
* What happens when vibe coding gets you to 80% but can’t get you the rest of the way
* Security risks that are easy to miss: exposed API keys, plain text credentials, and more
* HIPAA, PCI compliance, and other reasons to think carefully before vibe coding certain apps—no matter how cool they sound
* The “procrastination parades as productivity” trap, and how to tell if your tinkering is actually serving your business
* Why token costs matter more than most people are factoring in right now
* Joe’s lawnmower test: when just paying for the tool is the smarter move
* The 10-point framework Joe uses before starting any vibe coded project
"When you vibe code something and something goes wrong, it is a very helpless feeling if the AI can't fix it, because now you are taking something that you didn't write that you may not be familiar with. With my iOS app, it was like, 'There's something wrong with P list.' And I'm like, 'What is P list? Like, I don't know what that is. Can you just fix it?' And it's like, no. At some point you will hit this plateau where you can't take the app any further or something is broken. And you either just have to shrug or actually hire a developer to fix it." — Joe

About our Guest
Joe Casabona Website | LinkedIn
Joe's 10-point checklist for coding with AI
Connect with Us
Listen on Spotify
Listen on Apple Podcasts
Meg Casebolt
Jessica Lackey


This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit aggressivelyhuman.substack.com

The AI adoption conversation in early 2026 is loud, fast, and full of warnings that you’ll be left behind if you don’t move now. But for most of us, the real question isn’t how fast you can adopt every new tool—it’s whether the tool or use case is right for your business.
In this episode, Jessica shares her own journey from existential AI FOMO in January (OpenClaw! ClaudeCode! Opus! Oh my!) to a more grounded relationship with which tools actually serve her business. Meg brings her framework on early adopter cycles, the system stress that follows loud hype, and why the people yelling the loudest about a technology are usually trying to prove something—to themselves or to you.
We talk about what friction is actually telling you, the work AI is currently best poised to eliminate, and how to stay informed enough to make good decisions without getting sucked into the daily noise.
* Jessica’s emotional arc from existential AI FOMO in January 2026 to a calmer, more grounded approach—and what actually changed
* Why early adopters yelling about a technology often signals the window of advantage is already closing
* The “friction as signal” reframe: how slowing down through manual processes reveals whether you should be doing something at all
* What AI will realistically eliminate—repetitive, clerical, execution tasks—versus what stays irreplaceable (strategy, point of view, and looking someone in the eye)
* The “glue jobs” problem: roles that hold organizations together without driving revenue directly, and the risk of automating them away before you understand what they do
* How to tell the difference between adopting a tool because you genuinely need it versus keeping up
* Why understanding how a system breaks is more valuable than just using it to generate output—and how building things yourself teaches you that
* How to stay aware enough to make smart decisions without falling into constant FOMO-checking as a business strategy
"The friction slows me down enough to be like, 'This was a poor choice, Jessica. Strategically, I'm asking, am I dragging my feet because it's taking a long time? Or am I dragging my feet because this is not the right move? If you don't drag your feet, you never ask those questions. And I was partway through building out Five Foundations as a suite of courses and I stopped and said, wait—is this actually what I want to do? And the answer was no. The friction gave me the time to figure that out." -Jessica

Resources
TBM 417: Before You Fire All Your Glue People Because of AI
Matt Schumer “Something Big is Happening” Wikipedia page
Citrini’s 2028 Global Intelligence Crisis
The Panel with Justin Jackson and Brian Casel
Connect with Us
Listen on Spotify
Listen on Apple Podcasts
Meg Casebolt
Jessica Lackey


This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit aggressivelyhuman.substack.com

Business governance — LLC, C-Corp, Co-Op…. sounds boring as hell. But is it?
Instead of setting up a traditional partnership or agency, Anna and Janel of Strange Birds did something different — they spent a year building out the governance structure of a worker-owned cooperative.
Anna has been running Strange Birds, an idea specialist consultancy for over six years; Janel joined two and a half years ago after leaving Meta — and instead of just adding a second name to the partnership agreement, they restructured the whole business as a worker-owned co-op.
We dig into what a cooperative actually is and why it’s meaningfully different from a standard partnership, what the business case actually looks like (the survival stats are genuinely wild), and how going through the process forced them to have all the money, roles, and “what if someone stops pulling their weight” conversations that most business partners quietly avoid.
* What a worker-owned cooperative actually is — and how it’s legally different from a regular partnership or LLC
* Why 90%+ of co-ops outlast the 10-year mark while most traditional businesses fail within five
* The personal case for co-ops: why the creative services agency model is broken and what a democratic ownership structure actually fixes
* Equal pay, dependents, and the specific messy money conversation they had to have before finalizing their bylaws
* Why the cooperative legal structure took them a year to formalize — and why that’s a feature, not a bug
* How defining clear roles replaced the “we should all be interchangeable” model and made both partners more confident and effective
* What “humans first” looks like in practice when a kidney infection takes someone out for a month
* The structural safety nets they built into their bylaws for future members
* How practicing difficult internal conversations has made them better communicators with clients
* Keeping a 20-year friendship intact when you’re also co-running a business together — and the early warning sign when things start to feel tense
A cooperative model helps make sure we're all not just quote unquote horizontal — which, for people who have worked in horizontal business models, be real: was it actually horizontal? It was not. There was always a boss. Always a boss. There's always someone saying no. This is actually democratic. We all own it. We're all invested in it, and we all get paid equally regardless of if one person's services are bringing in more money — because a rising tide lifts truly all ships in a cooperative."

About our Guests
Strange Birds
Mentioned Resources
Lauren Edwards Flying the Coop episode
Incorruptible by Eric Ries
Connect with Us
Listen on Spotify
Listen on Apple Podcasts
Meg Casebolt
Jessica Lackey


This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit aggressivelyhuman.substack.com

AI clones and GPTs are increasingly common, which brings forth so many questions! But the “aggressively human” question we’re exploring today is how much care, curation, and legal scrutiny goes in before you release it to real people.
In this episode, we talk with return guest Diann Wingert, ADHD business coach and host of the ADHDish podcast, about Di AI — her newly launched digital coaching clone. Diann is a former psychotherapist turned business coach who works exclusively 1:1, and she just released a Coachvox-built AI version of herself in beta to current and previous clients.
We get into the technical, client experience, and ethical questions of cloning a coach. We cover what Diann chose to include (and exclude) in the training data, the months of fine-tuning behind the scenes, the legal work she did before launch, and why she’s not trying to turn this into a passive income stream.
* Why Diann has been wanting to “clone herself” since Dolly the sheep — and what finally made it possible
* Choosing Coachvox over other clone platforms (and why support and training mattered more than the tech)
* Why guest episodes and client success stories were excluded from the training data
* The painstaking fine-tuning process: ~100 questions per framework, batched over months
* Removing language she’d never use — like “ADHD is a superpower” and “work with your brain, not against it”
* The legal work behind the launch: hiring an AI-savvy attorney to rewrite her privacy policy and terms of use
* How she’s drawing the line between coaching and therapy inside the bot
* Why she’s giving Di AI away in beta — and what data she’s actually trying to gather
* The OCEAN personality score and how it’s reshaping how Diann talks about her ideal client
* Risk tolerance, exit strategy, and what it looks like to run an experiment in public

“I don’t want somebody going to the internet or to Gemini or Perplexity or Claude or ChatGPT when they’re running a business and they have an ADHD brain. I would much rather they go to something that is just my information. Not because I know everything about running a business with ADHD — ‘cause nobody knows everything — but because the information has already been filtered through 30 years of experience, multiple frameworks and models. And they can’t get an opinion at this point from Di AI that is not entirely consistent with that.” — Diann Wingert

About our Guest
Di AI
Coachvox
Website
Podcast
LinkedIn
OCEAN personality test
Connect with Us
Listen on Spotify
Listen on Apple Podcasts
Meg Casebolt
Jessica Lackey


This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit aggressivelyhuman.substack.com

We’re both mid-launch as we record—and neither of us is following the big launch playbook. In this episode, we turn the mic on our own businesses as we’re launching while life-ing. Meg renamed her core program from a membership to a mentorship—the Signal Mentorship—and is figuring out how to re-sell something she’s already sold many times over when the positioning has changed. Jessica sold her six-month cohort almost entirely through personal emails, membership upgrades, and alumni re-enrollments, and is calling it what it is: a cozy launch. (Meg has to remind Jessica that that still counts as a launch!)
We get into the behind-the-scenes logistics—pre-scheduling emails before vacation, hiring a copywriter if you’ll never actually write launch emails yourself, when and why to use paid workshops, and the feeling of “enoughness” in a launch.
* Meg’s rebrand from the Content Love Lab to the Signal Mentorship—and why “membership” wasn’t the right word anymore
* Pre-scheduling launch emails around vacation, jury duty, and client deadlines that don’t pause for you
* Jessica sent 50+ personalized emails instead of a launch sequence—and why she felt like it “didn’t count” as a launch
* The difference between “cozy launching” and “lazy launching” (hint: cozy launching is still a lot of work)
* Why Jessica doesn’t use bump offers, fast-action bonuses, or urgency deadlines
* How we invite discernment in our launches
* What happens when you’ve launched the same program seven times and your people already know the rhythm
* Meg’s paid challenge experiment that went sideways when people came back a month late demanding access
* The power of low-lift launches over time
* How authority compounds when you stop resetting and start building slowly over years
"I had a cozy launch, as [former guest] would say. And this episode is all about cozy launching. But it's because I have been launching the same program with the same time spots for three years now. March and September. It also really helps when you have something that people can re-up and enroll again, because I'm never starting from zero. I'm starting from that one person who said, yes, I'd like to do this again with you." —Jessica

Programs We’re Referencing
Signal Mentorship
Define Your Foundations (cohort re-launching in September)
Deeper Foundations Membership (open every day!)
Providers Jessica uses
Jessica’s Launch Copywriter Courtney Fanning
Bev Feldman Kit Email Advisor
Connect with Us
Listen on Spotify
Listen on Apple Podcasts
Meg Casebolt
Jessica Lackey


This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit aggressivelyhuman.substack.com