Page 2 / 6

The Observatory

The Wire
Retention Is the Bucket: What Compounds When Your Agent Never Forgets
Retention Is King · Andrew Chen
Why it matters
If your agent onboards users beautifully and they still drift away, you have not built a product — you have built a demo. Andrew Chen's math on retention curves is the first thing to internalize before you spend another dollar on growth.
Read the dispatch
Lesson · Distribution
Retention Is the Bucket Your Agents Fill Into
Acquisition without retention is a tax you pay forever — and in AI-native products, where agents generate output continuously, a leaky bucket drains faster than any human-paced product ever could.
Read the lesson
The Wire
When Your Agent Is the Experiment, De-Risk the Riskiest Assumption First
Running Lean · Ash Maurya
Why it matters
Ash Maurya's method was built for founders who can't afford to build the wrong thing — and now your agents are building faster than ever, which makes the wrong thing arrive faster too.
Read the dispatch
Lesson · Building & Agents
De-Risk the Agent, Not Just the Business Model
Running Lean by Ash Maurya argues that founders waste time building the wrong thing because they never ruthlessly surface their riskiest assumption first. That discipline translates directly to AI-native companies, where the riskiest assumption is often not the market but the agent itself.
Read the lesson
The Wire
Agents Ship Fast; the Messy Middle Still Belongs to You
Founders at Work · Jessica Livingston
Why it matters
Livingston's interviews prove that the founders who survived didn't have better plans — they had better reflexes when the plan fell apart. That instinct is exactly what you can't yet delegate to an agent.
Read the dispatch
Lesson · Foundations
What Embarrassing Beginnings Teach the Agent-Era Founder
The founders Jessica Livingston interviewed for Founders at Work did not execute clean visions — they pivoted, stumbled, and survived on stubbornness while building things that looked trivial from the outside. That pattern maps directly onto what it means to build an AI-native company today, where the temptation is to let agents handle the mess while you manage from a distance.
Read the lesson
The Wire
When Your Cap Table Has Agent Rows, Economics and Control Still Rule
Venture Deals · Brad Feld & Jason Mendelson
Why it matters
If you are raising money to build an AI-native company, the fundamental negotiation hasn't changed — but the story you're telling investors has, and that gap will cost you if you walk in unprepared.
Read the dispatch
Lesson · Human Judgment
What VCs Actually Control (And What You Must Keep)
The term sheet is not a formality — it is the moment economics and control are allocated, often permanently. Venture Deals by Brad Feld and Jason Mendelson makes this plain: strip every clause down to those two questions, and you understand the whole game. For founders building AI-native companies, the stakes are higher, because the control questions in a term sheet map directly onto the judgment questions inside your product.
Read the lesson
The Wire
Unscalable Work Is the Lab; Agents Make It Cheaper to Run
Do Things That Don’t Scale · Paul Graham
Why it matters
If your first hundred users don't feel personally obsessed over, you won't know what to automate — and right now, agents let you do that obsessing at a cost that has no precedent.
Read the dispatch
Lesson · Distribution
What Doesn't Scale Is What Teaches You
When software is cheap to build and agents can ship features overnight, the unscalable early work is not a phase to escape — it is where you learn what to actually build and for whom.
Read the lesson
The Wire
Hooks Were Designed for Humans. Now Your Agent Runs the Loop.
Hooked · Nir Eyal
Why it matters
When retention is automated, the ethical weight of the habit loop lands entirely on the founder's intentions — not the product's friction. That's the design problem you actually have.
Read the dispatch
The Wire
When the Agent Pipeline Breaks at 2 AM, You're Still the CEO
The Hard Thing About Hard Things · Ben Horowitz
Why it matters
Ben Horowitz's lessons about leading through chaos hit differently when your "staff" can spin up a hundred threads simultaneously — and fail across all of them at once.
Read the dispatch
Lesson · Building & Agents
The Habit Loop Has a New Host
In Hooked, Nir Eyal argues that the most durable products embed themselves through a repeating cycle — trigger, action, variable reward, investment — until return becomes automatic. For founders building with agents, the same loop applies, but the host of the habit is no longer just your interface; it is your entire agentic system, and trust is the variable reward that keeps users coming back.
Read the lesson
Lesson · Foundations
When There Is No Playbook, You Still Have to Ship
Ben Horowitz built The Hard Thing About Hard Things around a single uncomfortable truth: the moments that define a company have no recipe, and the founder who survives them is the one who manages their own psychology well enough to keep making decisions. That lesson lands differently now, because AI-native founders face a version of the same problem with a new layer of complexity — the system that does the work cannot manage its own psychology, so yours has to be steady enough for both of you.
Read the lesson
The Wire
Your Agents Won't Rise to Your Goals Either — They'll Fall to Their Systems
Atomic Habits · James Clear
Why it matters
If you're still tuning agent behavior by rewriting prompts after something breaks, you're goal-fixing when you should be system-building. The difference compounds, and not in your favor.
Read the dispatch
Lesson · Human Judgment
Habit Architecture for Founders Who Ship with Agents
When you build with agents, your results are no longer the sum of your individual decisions — they are the output of the systems you designed, the defaults you set, and the judgment boundaries you chose to hold or delegate. James Clear's Atomic Habits offers a translation key that founders building AI-native companies cannot afford to ignore.
Read the lesson
The Wire
Atomic Networks: The Minimum Viable Trust Unit for Agent Fleets
The Cold Start Problem · Andrew Chen
Why it matters
If your agents are doing the work, the network effects that make or break the product are being seeded right now — and most founders are too busy watching outputs to notice they're engineering a cold start into their architecture.
Read the dispatch
Lesson · Distribution
Atomic Networks Still Work — Now Your Agents Are Nodes
Andrew Chen's Cold Start Problem teaches that network products do not grow uniformly — they grow by seeding the smallest self-sustaining unit first, then expanding outward. For founders building AI-native companies today, that same logic applies: your agents, your workflows, and your early users form a network, and if the atomic version cannot sustain itself, scale will not save you.
Read the lesson
The Wire
When Your Agent Ships Overnight, You Still Need Someone Who Cares Why
Inspired · Marty Cagan
Why it matters
If agents can execute faster than your team can reflect, the question of what to build — and for whom — becomes the only question that actually matters. Get that wrong and you're just failing at speed.
Read the dispatch
Lesson · Building & Agents
What Cagan's Product Teams Teach Agent Builders
Marty Cagan's Inspired argues that real product work is discovery work — teams must continuously learn what is valuable, usable, feasible, and viable before and during delivery, not after. For founders building with agents today, that framing is more urgent, not less, because agents make it dangerously easy to ship output before you have discovered what outcome you actually need.
Read the lesson
The Wire
Validated Learning Still Runs the Loop — Now Agents Close It Faster
The Lean Startup · Eric Ries
Why it matters
If your agents are shipping output nobody measures, you're just burning compute on vanity. The build–measure–learn cycle Ries built for human teams gets sharper and more dangerous when the loop runs at machine speed.
Read the dispatch
Lesson · Foundations
Validated Learning When Your Prototype Costs an Afternoon
The Lean Startup by Eric Ries gave founders a discipline for operating under uncertainty: treat every product decision as an experiment, measure what actually changes behavior, and decide deliberately whether to pivot or persevere. That discipline matters more than ever when an agent can generate a working prototype in an afternoon — because speed without a feedback loop is just faster confusion.
Read the lesson
The Wire
When Your Agent Is Always System 1, You Need to Engineer System 2
Thinking, Fast and Slow · Daniel Kahneman
Why it matters
Kahneman's core finding is that fast, confident judgment is the default — and it's wrong in predictable ways. If your agents are making dozens of decisions a minute, you've just industrialized the failure modes he spent a career cataloging.
Read the dispatch
Lesson · Human Judgment
When Your Agent Is Overconfident and You Are Too
Kahneman's dual-system model is the missing mental model for founders who delegate consequential work to agents, because agents inherit the fast-thinking failure modes and humans rarely slow down enough to catch them.
Read the lesson

Load earlier observations

Showing 48 of 144

The Observatory

The observation log

Retention Is the Bucket: What Compounds When Your Agent Never Forgets

Retention Is the Bucket Your Agents Fill Into

When Your Agent Is the Experiment, De-Risk the Riskiest Assumption First

De-Risk the Agent, Not Just the Business Model

Agents Ship Fast; the Messy Middle Still Belongs to You

What Embarrassing Beginnings Teach the Agent-Era Founder

When Your Cap Table Has Agent Rows, Economics and Control Still Rule

What VCs Actually Control (And What You Must Keep)

Unscalable Work Is the Lab; Agents Make It Cheaper to Run

What Doesn't Scale Is What Teaches You

Hooks Were Designed for Humans. Now Your Agent Runs the Loop.

When the Agent Pipeline Breaks at 2 AM, You're Still the CEO

The Habit Loop Has a New Host

When There Is No Playbook, You Still Have to Ship

Your Agents Won't Rise to Your Goals Either — They'll Fall to Their Systems

Habit Architecture for Founders Who Ship with Agents

Atomic Networks: The Minimum Viable Trust Unit for Agent Fleets

Atomic Networks Still Work — Now Your Agents Are Nodes

When Your Agent Ships Overnight, You Still Need Someone Who Cares Why

What Cagan's Product Teams Teach Agent Builders

Validated Learning Still Runs the Loop — Now Agents Close It Faster

Validated Learning When Your Prototype Costs an Afternoon

When Your Agent Is Always System 1, You Need to Engineer System 2

When Your Agent Is Overconfident and You Are Too