Optimize For Uncomfortable Conversations

On Saturday I shared a deep dive on how creating a category is overrated. Join the community to get access to my library of 45+ deep dives, and a new one in your inbox each Saturday.

Today at a glance:

💡 Opportunity: CAPTCHA 2.0

Every time I use CAPTCHA I’m convinced there has to be a better solution. And pretty soon there will need to be, as it seems like AI systems have already “solved” existing CAPTCHA solutions. Makes you wonder if Elon is right about his approach to fighting bots.

Anyway, given how prevalent CAPTCHA has become this presents a huge opportunity to create whatever comes after it.

This issue is that CAPTCHA 2.0 can’t just be a “better” or “harder” CAPTCHA puzzle. AI will figure it out much faster than humans at this point.

Instead, to come up with the next solution we need to return to first principles.

Controversial as it’s been so far, Worldcoin would solve this. By proving that the entity solving the CAPTCHA puzzle, or logging in (because you wouldn’t even need the puzzles in this case), was a human rather than a bot Worldcoin would effectively block bots.

There are likely some intermediary steps to sufficiently decentralize CAPTCHAs until Worldcoin reaches prominence (creepiness concerns aside).

🧠 Framework: 11 Star Experience

In the early days of Airbnb, co-founder Brian Chesky would consistently ask users “what can we do to surprise you? What can we do, not to make this better, but to make you tell everyone about it?”

If you’ve been a subscriber for a while you know I’m not a huge of NPS (though it does have its uses), and Brian created his own framework to understand the experience his users were having on the platform instead of just using NPS as well.

He calls it the 11 star experience framework:

It sets a benchmark for what each level of user experience is like within your product that you can refer back to over time.

Trying to prioritize new features? Consider how much each one moves your users closer to an 11 star experience.

To really nail this you need to be deeply in tune with your users — you need to understand not only their wants and needs, but also their expectations.

Most importantly, what are their non-negotiables and what are the things they themselves likely can’t even think of — but would blow their minds?

📈 Trend: AI Capability Control

Ah yes, the AI safety debate.

The struggle between making sure AI systems are aligned to human interests, or trusting that they either will be or they won’t pose a threat to us regardless.

One of the terms that’s gotten more attention recently is “AI Capability Control” which is basically exactly what it sounds like.

The argument for “safe" and “ethical” AI is a tricky one — who’s defining the ethics and acceptable behavior? It sounds like an east way to introduce bias into a system to me, intentionally or otherwise.

But even Elon Musk, who says xAI is working on a “truth seeking” AGI called TruthGPT, advocates for some level of AI safety.

💬 Quote: Optimize for Uncomfortable Conversations

Honest truth: the universe’s (and society’s) nature is uncontrollable cha. We sometimes delude ourselves into thinking we can control it… but why?

The moments that change the trajectory of our lives and shape our thinking are the ones we least expect, that are unplanned, surprising, and sometimes filled with tension.

This is also true with your work as a founder.

Do you already know what will happen in a meeting? Does the team already agree on it? Great, cancel the meeting.

Are there discussions you’re not having because they’re uncomfortable, or make apparent risks that you’ve been shielding your team from? Set a meeting for that instead.

The only way we grow as founders — as leaders — is to embrace these moments.

🔗 Houck’s Picks

• 16 point checklist for what a great startup idea needs to hit (Link)

• The advice Alexis Ohanian gives most often (Link)

• Good case studies for solving the cold-start problem (Link)

• The one thing the best founders do in common (Link)

• Netflix co-founder Marc Randolph on bad ideas (Link)

💡 How I Can Help