Dec 18, 2024 · 42m · startup-ideas
The framework for building AI startups from a product design genius
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Product leader and entrepreneur Mike Hudack joins Greg Isenberg to unpack essential frameworks for building AI startups, value-based pricing models, B2B versus B2C dynamics, and the organic distribution strategies behind his fintech venture, Sling Money.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Greg holds 33.4% of the talking time here. How this is scored →
Greg as informed peer 4.6 Guest teaching 4.8 Guest disagreement 1.2 Greg pushing back 0.7
speaking balance: gold is Greg, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 25:31 Rejecting the anti-copying dogma
Hudack explicitly rejects the common product taboo against copying, arguing that refusing to adopt proven primitives is foolish.
Hardest push from Greg ▶ 36:09 Questioning price increase dogmasGreg prompts Hudack with the traditional startup axiom that products can only lower prices rather than raise them.
Biggest teaching moment ▶ 37:20 Monzo pricing playbook lessonHudack educates Greg on how Monzo successfully raised fees on ATM withdrawals and card replacements without alienating loyal users.
Greg holds their own ▶ 2:59 Greg shares structured AI agent thesisGreg demonstrates his product expertise by detailing a four-point framework for building high-utility, low-friction AI agents.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Greg as informed peer | Guest teaching | Guest disagreement | Greg pushing back | Why |
|---|---|---|---|---|---|---|
| Designing Frictionless AI Agents and Everyday Workflows | 6 | 3 | 1 | 1 | Greg lays out his own four-part framework for building valuable AI agents, which Hudack validates and builds upon with personal product experiences. | |
| Sponsor Break: Accelerating Founders with Startup Empire | 4 | 6 | 1 | 0 | Hudack introduces several portfolio startups like Greenlite and Granola, walking Greg through their specific enterprise compliance and note-taking mechanics while Greg pulls them up on screen. | |
| Voice Automation and Niche Opportunities with Bland AI | 6 | 3 | 1 | 1 | Greg demos Bland AI and proposes geographical arbitrage opportunities, prompting Hudack to share a complementary example of AI-driven debt collection. | |
| Comparing B2B and B2C Startup Pathways | 4 | 5 | 1 | 1 | Hudack breaks down the mathematical trade-offs between B2B and B2C dynamics, while Greg chimes in with founder psychology regarding consumer failures. | |
| The Regret Minimization Framework for Choosing Startup Ideas | 4 | 5 | 1 | 0 | Hudack walks through a regret-minimization decision matrix for launching Sling Money, which Greg connects back to his own venture experience. | |
| Sponsor Break: Scaling Organic Traffic with BoringMarketing | 4 | 5 | 2 | 1 | Hudack challenges the purist notion against copying features, reframing foundational ideas as primitives that every company should adopt, while Greg agrees on maintaining core balance. | |
| Evolving Distribution: Organic Native Content and Building in Public | 5 | 4 | 1 | 1 | Hudack details modern organic distribution and community-led launches in Kenya, while Greg adds insights on reverse-engineering short-form video hooks. | |
| Value-Based Pricing and Raising Rates for AI Products | 5 | 6 | 2 | 1 | Hudack rejects conventional wisdom that companies cannot raise prices post-launch, citing his Monzo experience, while Greg reinforces the concept using Uber's market expansion. | |
| Analyzing Mark Zuckerberg and Meta's Open AI Strategy | 3 | 6 | 1 | 0 | Hudack provides an insider perspective on Mark Zuckerberg's open-source AI strategy as a calculated move to commoditize model layers and leverage proprietary training data. |
Statements from this episode (16)
Isenberg: Build AI agents by adding LLM understanding to one high-impact action
“How do you build a smart agent that is going to be valuable? Start with one action that matters. Add LLM understanding. Build tight feedback loops, learn from every interaction, and expand slowly.”
Isenberg: Manual niche service agencies are prime targets for AI disruption
“Agencies are a really good business right now because it's super manual, like it's humans doing things. It's in a particular niche, like a good agency is focused on SaaS, FinTech. They're the best at, you know, just this one niche.”
Hudack: Gradient AI dramatically outperforms Intercom across all shared customers
“These guys are dramatically outperforming Intercom in all of the, with all the customers that they have.”
Isenberg: Huge opportunity in running localized, country-specific arms of AI startups
“And I think that there's a huge opportunity in taking some of these businesses, partnering with them and being like, I'll just do your, you know, Canadian arm.
And, you know, people think, oh, that's so small.
It's such a small way of thinking.”
Isenberg: Canadian fintech Wealthsimple is still valued at $3.5 billion
“You know, well, simple, which is the Robin Hood of Canada is still like a three and a half billion dollar company, you know, there's opportunity in geographies.”
Hudack invested in YC AI debt collection startup despite initial skepticism
“There's another one I invested in recently.
It's a YC company that is doing debt, AI debt collection, which I would have thought was like a terrible idea.
I can't think of the name off the top of my head, but the way the guy explained it, which I thought was r…”
Hudack: Founders must build consumer products to create a $100B company
“You're also probably not going to create like a hundred billion dollar company in doing that, you know? If you want to do that, you got to go consumer, I think.”
Hudack: Sling Money went viral in Kenya after 2.5 years of building
“We're trying to build like WhatsApp for money or a global Venmo. You know, we've been building for two and a half years. It's super difficult. You know, you iterate, you keep building, and then, you know, maybe you have a moment. We had a moment last week. We …”
Hudack: Founders should replicate common product primitives instead of inventing everything
“I think there are just certain primitives or certain ideas that become commonplace and, like, should be replicated by everyone. And you have to be, like, you know, it's, like, impossible that you're gonna come up with all of those yourself.”
Hudack: Targeted Facebook ads no longer work well for acquiring early adopters
“I think that there was probably a golden age in which the answer would have been like,
You know, target exactly the right message at exactly the right people on Facebook.
Find the, like, 3400 people who are your early adopters and just, like, get it in front o…”
Hudack: Organic distribution requires adopting platform-native communication formats
“If you want to target like techie people, you probably want to do it on Twitter X. If you want to target
Sales people, you probably want to do it on LinkedIn.
You know, like, it just, you know, and there's a format, and there's a way of thinking, and a way of …”
Hudack: Price AI by heavily discounting total labor and emotional damage saved
“With an AI company, it feels really clear to me that you look at the number, the amount of human input that you're saving, And the cost of that human input, and I mean total cost, I mean salary, emotional damage that you have to sustain in doing the work, you …”
Hudack: Monzo's top pre-COVID revenue stream was overseas debit interchange fees
“Our number one revenue stream was actually people using their debit cards overseas, and the interchange you get from that, and that went away overnight, and we were totally screwed.”
Hudack: Startups can raise prices later on products that customers truly love
“If you build a thing, which is like, Beautiful and people love and has a place in their lives. You can charge more for it later. And you know, that goes against conventional wisdom. And I think if you have that knowledge it's a secret weapon because it's easie…”
Hudack: Meta's open-source strategy smartly commoditizes competitors' closed AI models
“I think that the strategy that they have right now of like open sourcing a bunch of models, you know, makes a ton of sense. It's like the right strategic move. If you're in their position you're kind of like salting the earth in a pretty meaningful way, you kn…”