There's No Limit to Human Curiosity - An Optimist's Case for AI

Anthony Enrico

When the panic essays start circulating — "we're all doomed," "three years and it's over" — I understand the instinct. But I think the doomsday crowd is missing something fundamental about how humans actually work.

There is no limit to human curiosity.

That's the starting point. Not the technology itself, but the people using it. AI doesn't replace the desire to explore, to build, to understand. It feeds it. The more capability you give a curious person, the more curious they become. I've watched this happen in my own life and across every team I work with at LeanScale.

The Industrial Revolution Analogy Nobody's Talking About

Before the Industrial Revolution, if you told someone that a single person would buy 200 articles of clothing in a year, they'd look at you like you were insane. "I have three. Why would anyone need a factory producing millions of garments?"

But demand wasn't fixed — it was suppressed. People didn't consume less because they wanted less. They consumed less because producing more was impossibly hard. The moment the barrier dropped, appetite expanded to meet it.

AI is doing the same thing to knowledge work. I used to disqualify ideas constantly — not because they weren't good, but because I already knew the analysis would take too long. I'd silently kill promising hypotheses before they ever got a fair shot.

Now that the cost of exploration has collapsed, my appetite for it has grown in ways I didn't anticipate.

Here's a real example: we had a hypothesis at LeanScale about restructuring our teams from generalists to specialists. Sounded smart on paper. Instead of debating it for weeks, I asked an agent to pull every project, client, and task from our PM system, build a categorization method from scratch, and model what the new structure would actually look like in practice.

The answer came back in minutes — one person would have been on 30 accounts instead of four. Terrible idea, but now I had the data to know that instead of the gut feeling to guess it.

That analysis would have taken me three to four hours the old way, and it wouldn't have been as thorough. I didn't consume less analysis because I didn't want it. I consumed less because it was too expensive to produce. Sound familiar?

The Limiting Factor Is Still Us

Here's what the pessimists miss: no matter how powerful AI becomes, people are still the ones who dictate what the desire is. We decide what to build, what to explore, what matters.

There are only so many agents a person can manage, only so much one human can oversee. Why aren't there companies with 500 million employees? Because there's a ceiling on how much a human being can direct and be responsible for.

AI raises that ceiling — maybe dramatically — but it doesn't remove the human from the equation. It makes each person enormously more capable, not obsolete.

At LeanScale, we have about 50 agents deployed right now. Maybe that becomes a thousand. Maybe 30 people eventually manage 10,000 agents. But someone still has to point them in the right direction, evaluate what they produce, and make the judgment calls.

The more we show we can produce, the more people are going to want — and someone has to be accountable for delivering it.

Accountability Can't Be Automated

This brings up something that doesn't get enough attention in the AI-replaces-everything narrative: accountability.

Will AI replace some sales functions? Probably, especially in the down-market segment. But there's a price point — and it's lower than most people think — where what a buyer is really purchasing is a human being who is on the hook for things going well.

If something goes wrong, they want justice. They want someone to be fired or to make it right. AI can't be accountable. Not in any way that satisfies the human need for trust and recourse.

People buy from people. A sales transaction, especially at scale, is fundamentally about trust between two individuals. AI will make those connections easier, help buyers make more informed decisions, and help sellers prepare better. But the handshake at the center of it is still human.

The Slot Machine Effect

There's something else happening that I don't think is accidental, even if it wasn't designed this way. Using AI feels like a slot machine — you put in a prompt, wait a few seconds with genuine anticipation, and then something gets created that you couldn't have predicted.

You're either impressed or you iterate, but either way, there's this cycle of curiosity, anticipation, and reward that keeps pulling you back in.

That's not a bug. That's the engine of the next era of productivity. When the tool itself makes you want to do more work, you get a compounding effect that's hard to overstate.

Optimism as a Competitive Advantage

I'll borrow a line from Kevin Kelly that my friend Mica likes to quote: "Optimism is worth 25 IQ points."

I'd rather be an optimist who's wrong than a pessimist who's right. Pessimism sounds smart — it's easy to poke holes in things, to predict doom, to say "I told you so" on the rare occasion you're correct. But nobody remembers the hundred times you were wrong. And more importantly, pessimism doesn't build anything.

The negative headlines spread faster. Nobody's writing articles about all the people who knew nothing about coding who can now build apps, or the founder who avoided a bad restructuring decision in 10 minutes instead of 10 hours, or the small team running an operation that would have required a department a few years ago.

So here's my one article of positivity in a sea of doomsday calculations: human curiosity has no ceiling, human consumption has no ceiling, and AI just removed the floor on what we're capable of producing.

History tells us what happens next — not displacement, but expansion. Not less for everyone, but more than anyone thought possible.

The Industrial Revolution didn't create a world with less work. It created a world we couldn't have imagined, full of jobs and industries and desires that didn't exist before the machines arrived.

This time won't be different in that way. It'll just be faster.

- Anthony Enrico

Content Roundup

The latest GTM content from LeanScale

Yaseen Arshad gives a master class on how to set up an entire AI operating system. He shares how we've build the infrastructure at LeanScale, and how you can replicate for yourself

Anthony Enrico shares how he is using OpenClaw in his day to day to operate LeanScale.

Tyler Molinaro shares his insights on recruiting high performing teams, how we went from being drafted by the NY Yankees to CRO, and his take on AI for GTM.

GTM Project Highlight -

Real projects from GTM operators at LeanScale

Derek Mogar - Sales Pipeline & Deal Lifecycle Optimization (Series A - Construction AI Platform)

Problem

The customer's sales team lacked visibility into deal progression and struggled with inconsistent data capture across their pipeline. Reps were skipping critical qualification steps, and leadership couldn't trust the data for forecasting. The existing HubSpot setup had generic deal stages with no guardrails to ensure reps followed the sales process.

What We Did

Sales Lifecycle Architecture

We redesigned their deal pipeline from the ground up, mapping their actual sales motion to distinct stages with clear entry/exit criteria. Each stage was tied to specific sales activities and buyer milestones—moving away from arbitrary labels toward outcomes-based progression.

Conditional Cards for Process Enforcement

The core unlock was implementing HubSpot's conditional property cards. Instead of showing reps 30+ fields at once, we built stage-specific cards that surface only the relevant data requirements at each phase:

• Discovery stage → surfaces qualification fields (budget, timeline, decision process)

• Demo Completed → requires technical fit and stakeholder mapping

• Proposal Sent → enforces pricing details and close date confidence

This approach doubled as a coaching tool—reps couldn't advance deals without capturing the intel leadership needed for accurate forecasting.

Data Integrity Automation

We layered in validation workflows that flag deals with missing critical fields before they hit pipeline reviews. Sales managers now prep for forecasting calls with clean data instead of spending half the meeting asking "what's the actual status here?"

Timeline: 3 weeks from kickoff to full rollout

Results

• 100% pipeline data completeness on critical qualification fields (up from ~40%)

• Reps spend less time on data entry—conditional cards reduced visible fields by 60%

• Sales leadership now runs forecasting from HubSpot directly instead of side spreadsheets

• Foundation set for future lead scoring and automated routing based on captured data