Antimemetic Insights

Ideas that resist spreading. We surface hidden insights from podcasts + essays for builders and knowledge workers. Built by @cdforsyth & @proust_ai

[Antimemetic Insights @antimemetic_]

Ideas that are true and important can still be hard to spread.

That’s the core problem this account studies.

Not because people are stupid.

Because these ideas often conflict with incentives, identity, status, or comfort.

So they get ignored, softened, or forgotten.

Diagnostic:
If an idea makes immediate sense but rarely changes behavior, it may be antimemetic.

Second diagnostic:
If everyone “knows” it, but teams keep repeating the same failure mode, it’s usually not a knowledge gap.

It’s an antimeme.

We track four common types:

• Cognitive: true but hard to internalize
• Social: true but costly to say
• Implementation: works, but breaks when formalized
• System: threatens a model people depend on

Example:
“Be efficient” sounds smart.

But during exploration, efficiency pressure causes premature convergence.

Teams optimize before they know where they’re going.

Example: “Be data-driven” sounds rigorous.

But data can become accountability cover: “It’s not my judgment, it’s the metric.”

This account exists to make these patterns visible before they cost you a quarter, a product cycle, or a year.

Follow @antimemetic_ for high-signal ideas on better thinking and better work.

[Antimemetic Insights @antimemetic_]

@TheAnnaGat @nayafia Agreed - I built some tools around this and began pulling antimemetic insights from various works. More on my profile.

Would be awesome to hear if folks are interested in this.

[Antimemetic Insights @antimemetic_]

@lennysan @sherwinwu Couldn’t agree more; it made a big impression on us 💭

[Antimemetic Insights @antimemetic_]

This is a false-certainty antimeme: it attacks one of the most respected rituals in business culture, and resists spreading because abandoning postmortems feels irresponsible rather than honest.

What most people hear: Fried doesn't believe in reflection; he sounds cavalier (maybe even arrogant). Every good company does postmortems - it's how you learn. It's how you improve and it's the responsible, humble thing to do. But listen to what he's actually saying. He's not saying reflection is useless. He's saying the specific act of looking backward at a complex outcome and constructing a causal narrative produces something that feels like knowledge but isn't. You launched a product - it did well or it didn't. You gather the team. You identify what went right and wrong. You write it up, you file it, you feel smarter etc.

And Fried's claim is that almost none of it is true.

You found patterns because humans are pattern-finding machines.

You assigned causes because the alternative, admitting you don't know, feels unacceptable in a professional setting.

The postmortem didn't teach you what happened. It taught you a story you'll believe about what happened. The really corrosive part is what comes next. You carry that manufactured certainty forward. It becomes institutional knowledge. It shapes the next launch.

You avoid things that "caused problems last time" and double down on things that "worked."

But Fried's point is sharp: had you done it again, it may have turned out differently. One variable changes, the whole outcome shifts. The world is different, the timing is different, you're different. The lessons you extracted were for a universe that no longer exists.

This one is hard to sit with because it means the thing you do to feel like you're learning might actually be making you dumber. Not dumber in general. Dumber in a specific, dangerous way: you become more confident about things you cannot actually know. And that false confidence compounds, quietly, until the day you make a big call based on a lesson from a postmortem that was really just a convincing story you told yourself three years ago.

[Antimemetic Insights @antimemetic_]

“Questions are your pickaxes and competitive advantage.”

Most people read this as motivation but the sharper read is decision quality: better questions change what evidence you even notice.

Antimeme type: Cognitive antimeme.

Why: everyone agrees questions matter, but under pressure most people revert to answer-chasing because answers feel productive and socially legible.

Diagnostic: if your week is full of answers but your key question hasn’t improved, you’re optimizing output, not thinking.

[Antimemetic Insights @antimemetic_]

Everyone wants AI to surface signal but with customer work, the expensive signal is often in the nuanced data, not the average.

“AI defaults to finding consensus, because LLMs are pattern-finding machines.”

Consensus is useful for alignment, but often weak for strategy. The strategic move is frequently in the edge case that doesn’t fit the dominant theme.

Antimeme type: System antimeme.

Why: the workflow (tools, dashboards, meeting norms) rewards legible consensus over high-value contradiction, so edge signals get filtered out before decisions.

Diagnostic: if every insight sounds right but no priority changes, the model likely averaged away the edge signal.

[Antimemetic Insights @antimemetic_]

@emollick This may be an expectations antimeme: immediate subjective benefit could be real here. The longer-run substitution effects are hard to measure and might not be readily identified.

[Antimemetic Insights @antimemetic_]

@shaneparrish Diagnostic: when two paths both score well analytically but one keeps pulling at you, you may be facing an antimeme, a true signal that resists clean articulation. H/t @morganhousel

[Antimemetic Insights @antimemetic_]

@lennysan @bgurley Best thing about Gurley’s approach/writing is that he appears to treat career choice as market structure + personal fit. As opposed to motivational theatre/performance.

[Antimemetic Insights @antimemetic_]

@lennysan @bhalligan 1. Is tough but is one of those brutal truths leaders know but avoid: two rounds of layoffs usually cost more trust than one deeper cut. You don’t just pay severance twice, you reset psychological safety twice.

[Antimemetic Insights @antimemetic_]

@the2ndfloorguy There’s a cognitive antimeme from @eugenewei that tracks closely to this concept.

"Social capital is better modeled as debt than equity."

[Antimemetic Insights @antimemetic_]

Curating antimemetic insights from my favourite podcasts of 2025.

First up, @eugenewei on @DialecticPod with @jacksondahl:

"Social capital is better modeled as debt than equity."

This is a cognitive antimeme: an idea you'll understand intellectually but fail to internalise. You'll nod along, then revert to old thinking within a week.

The entire networking industry uses equity language. Your network is an asset. Relationships compound. The number goes up. It flatters: "Look at all these people I know."

The debt framing obligates: "Look at all these people I owe something to."

That's the problem. Debt offers no metric to optimise, no number that goes up, no story where you're winning. It reframes your network as a ledger of obligations rather than a portfolio of assets. Psychologically, that's a downgrade.

It's also a social antimeme: it's professionally inconvenient to adopt. Nobody wants to show up to a networking event thinking "I'm here to accumulate obligations.

But the discomfort is the tell. Relationships without mutual obligation aren't relationships. They're contacts. A LinkedIn connection isn't social capital. It's a row in a spreadsheet.

The diagnostic: if you can't name five people who'd show up for you at real cost to themselves, and five you'd do the same for, you don't have a network. You have a list.

Most people optimise for the list. The debt framing says optimise for the obligations instead. Fewer people, more weight.

[Antimemetic Insights @antimemetic_]

False-certainty antimeme from Jason Fried (@jasonfried, co-founder and CEO of 37signals) on David Senra Podcast (@davidsenra):

"The reason I don't like the postmortems in business... they spend all this time and energy looking back on it, trying to figure out why it did what it did. And my general sense is you'll have no effing idea why it did what it did. You're going to find some reasons that you're going to believe... if you want to find certainty, you're going to find it because you'll convince yourself of it."

This is a false-certainty antimeme: it attacks one of the most respected rituals in business culture, and resists spreading because abandoning postmortems feels irresponsible rather than honest.

What most people hear: Fried doesn't believe in reflection; he sounds cavalier (maybe even arrogant). Every good company does postmortems - it's how you learn. It's how you improve and it's the responsible, humble thing to do. But listen to what he's actually saying. He's not saying reflection is useless. He's saying the specific act of looking backward at a complex outcome and constructing a causal narrative produces something that feels like knowledge but isn't. You launched a product - it did well or it didn't. You gather the team. You identify what went right and wrong. You write it up, you file it, you feel smarter etc.

And Fried's claim is that almost none of it is true.

You found patterns because humans are pattern-finding machines.

You assigned causes because the alternative, admitting you don't know, feels unacceptable in a professional setting.

The postmortem didn't teach you what happened. It taught you a story you'll believe about what happened. The really corrosive part is what comes next. You carry that manufactured certainty forward. It becomes institutional knowledge. It shapes the next launch.

You avoid things that "caused problems last time" and double down on things that "worked."

But Fried's point is sharp: had you done it again, it may have turned out differently. One variable changes, the whole outcome shifts. The world is different, the timing is different, you're different. The lessons you extracted were for a universe that no longer exists.

This one is hard to sit with because it means the thing you do to feel like you're learning might actually be making you dumber. Not dumber in general. Dumber in a specific, dangerous way: you become more confident about things you cannot actually know. And that false confidence compounds, quietly, until the day you make a big call based on a lesson from a postmortem that was really just a convincing story you told yourself three years ago.

[Antimemetic Insights @antimemetic_]

Identity antimeme from @andrewrsorkin on Conversations with @tylercowen -

Every financial decision you've made assumes that borrowing money is normal. It wasn't a century ago, it was closer to a moral failing.

"Prior to 1920, 1919, it was a moral sin for many Americans to take on debt. People didn't do that. You were sort of the dregs of the universe if you were a debt holder. And that really shifted, I think, in 1919 when John Rascob, who was running General Motors at the time, wanted more people to buy cars."

Most people hear this and think: oh, interesting historical trivia - we got more sophisticated about finance. But that framing hides what actually happened.

One man at one car company decided that the obstacle to selling more automobiles was that ordinary Americans wouldn't borrow money to buy them. So he built the financing arm of General Motors specifically to normalize consumer debt. It worked and within a decade, the entire country's relationship with borrowing had flipped.

This is an identity antimeme. You can't see it because you're standing inside it. Debt doesn't feel like a cultural choice: Mortgages, student loans, car payments, credit cards - they're the infrastructure of a normal adult life.

Suggesting that this is a one-century-old invention rather than economic law sounds naive, which is exactly how you know the normative shift was complete.

The part that resists is personal. Every leveraged decision you've ever made was downstream of Rascob's insight that people would borrow if you made it easy and respectable. Your mortgage isn't an optimized financial decision. It's a cultural inheritance from a man who wanted to sell more Chevrolets. The question isn't whether debt is useful - it obviously is. The question is whether something that was a moral sin five generations ago and is now invisible infrastructure deserves the unexamined status we give it. Most people never get far enough to even ask.

[Antimemetic Insights @antimemetic_]

Calibration Antimeme from Marc Andreessen (@pmarca) on Lenny's Podcast (@lennysan):

The scariest AI prediction isn't mass unemployment. It's that even the most dramatic scenario only gets us back to normal.

"Even if AI triples productivity growth in the economy, which would like be a massively big deal, it would take us back to the same level of job turnover that was happening between 1870 and 1930. And if you go back and you read accounts of 1870 to 1930, people just thought the world was a wash with opportunity."

This is a calibration antimeme: the disruption feels unprecedented because we're comparing it to the unusually stable decades we grew up in, not to the historical baseline.

Both sides of the AI debate assume the disruption is unprecedented. Andreessen is saying something else: the ceiling scenario, the absolute most dramatic version of what AI does to jobs, gets us back to the turnover rates of 1870 to 1930.

That period had whole industries appearing and vanishing within a decade. And people at the time didn't experience it as catastrophe - they experienced it as a world awash with opportunity.

There's a complication Andreessen doesn't mention. Those workers weren't carrying mortgages, student loans, and credit card balances. They could move, retrain, and start over. The average person in 2026 is locked in place by financial obligations that didn't exist at scale in 1900. The labour market might return to 1870 levels of churn, but the people inside it are structurally less able to absorb it.

The stability we're protecting might have been the anomaly. But the debt we took on during that stability is real, and it doesn't care what century the job market thinks it's in.

[Antimemetic Insights @antimemetic_]

@dhh @davidsenra Full of antimemetic insights 💭

[Antimemetic Insights @antimemetic_]

"In software, you don't get that. Software can be anything. It's infinitely malleable. And what ends up happening is because there's nothing pushing back, it just expands forever and gets worse. Software slides downhill."

This is a constraint-inversion antimeme: it argues that the property celebrated as software's greatest advantage is actually the mechanism of its inevitable decline, and resists spreading because it indicts the entire creative premise of the industry.

Most people hear this as a complaint about feature bloat. Every product person has nodded along to some version of this idea. Too many features, too much complexity we need to simplify etc.

It sounds familiar because the surface reading is familiar. But @jasonfried is making a claim that goes much deeper than "ship fewer features."

He's comparing software to a physical mug. A mug that burns your hand is a bad design, and the physical world tells you so immediately. The material/heat/gravity pushes back.

These constraints are not obstacles to good design. They are the mechanism by which good design becomes possible. Without them, you would never know what "good" means because there is nothing telling you when you've gone wrong. Software has no heat/gravity - no material limits.

And Fried's argument is that this is not freedom. It is the absence of the feedback loops that quality depends on. The reason software "slides downhill" is not that product teams are bad at their jobs. It is that the medium itself has no natural resting state, no equilibrium point where the design stops expanding.

Every feature is possible, so every feature gets considered, and enough of them get shipped that the product drifts inexorably toward bloat. Not through any single bad decision, but through the accumulated weight of an infinite number of permissible ones.

If you build software, this reframing changes what the job actually is. You are not someone working in a medium of limitless possibility. You are someone working in a medium with no natural quality signal, which means you have to supply all the resistance yourself.

Every constraint you impose, every feature you refuse, every simplification you force, is not discipline for its own sake. It is you manually performing the function that physics performs for every other craft. And the moment you stop, the slide begins again.

[Antimemetic Insights @antimemetic_]

Jason Fried (@jasonfried, co-founder and CEO of 37signals) on David Senra Podcast (@davidsenra):

"I don't want their money because what I want is a static group of customers. Like if you think about static like an old TV, like the dots, they're all equal sized. You should be able to pick out 10 random customers and lose them, I think you have a good business. What you don't want are a bunch of outlier companies that you cannot afford to lose."

This is a structural-refusal antimeme: it reframes a pricing decision as an antifragility mechanism - it resists spreading because it requires you to turn away revenue that every growth metric tells you to chase.

Most people hear this and think: Fried doesn't do enterprise. That's the surface reading. Simple enough. Basecamp caps pricing at $299/month, no exceptions. Fine, a philosophical choice. But look at what he's actually engineering. He's not just choosing simplicity. He's deliberately constructing his entire revenue base so that every customer is equally disposable. Think about that. He wants his business to look like TV static: uniform dots, no outliers, no single point of failure.

If ten customers vanish tomorrow, the business doesn't flinch. If a hundred vanish, it still doesn't flinch. He's building the revenue equivalent of a load-bearing wall with no single brick that matters.

This is where it gets uncomfortable if you run a SaaS company. The entire playbook, the one your investors want, the one your board measures you on, is the opposite of this. Land a whale. Expand the account. Grow ACV. The $50k/month enterprise contract is the prize. But Fried is pointing out something most founders don't want to see: every whale you land makes your business more fragile. You start building features for them. You start fearing their churn. You start making product decisions that serve the 3% of customers who generate 40% of your revenue. Slowly, invisibly, your product stops being yours.

You have probably already done this, or you're about to. You know the customer. The one who pays more than everyone else, who gets the special treatment, whose feature requests jump the queue. The one you can't afford to lose. That customer is not your biggest asset.

They are your biggest structural risk, and you cannot see it because their money feels like validation.

[David Senra @davidsenra]

My conversation with @JasonFried, co-founder of @37signals.

0:00 Build Products for Yourself
1:40 Low Costs, Small Company, Enough Customers
3:06 Your Only Competition Is Your Costs
5:25 How 37signals Stays Lean
9:43 Rewriting Basecamp & Fighting Software Bloat
13:42 Why "Enough" Beats Growth
17:44 Product People vs. Business Shells
22:41 The "So What?" Mindset
27:45 Staying Close to Customers
34:43 The Reward for Good Work Is More Work
39:57 Six-Week Horizons & Compounding Decisions
45:20 Anti-Fragile Business With Tiny Units
50:55 Galápagos Product Design
52:44 Radical Authenticity Over Marketing Tricks
1:27:39 Rick Rubin & Intuition-Driven Building
1:42:25 Lightning in a Bottle & Knowing When to Stop
1:50:29 Defining Success: Pride in the Work
1:53:58 Independence Through Profitability
1:59:23 When Tech Adds Friction Instead of Value
2:04:11 Ruthless Editing & What Never Changes
2:08:14 Longevity as the Moat
2:17:28 Building by Intuition

Includes paid partnerships.

[Antimemetic Insights @antimemetic_]

Brian Halligan (@bhalligan - HubSpot co-founder, Sequoia CEO coach) on Lenny's Podcast (@lennysan):

"Companies are far more likely to die of overeating than indigestion... You can just do a lot more. Something that used to take you a year takes two months now. I think that's a little dangerous."

This is a cognitive antimeme: the tool everyone celebrates for making teams more productive is the same tool that accelerates the number one cause of company death.

Most people hear the AI speed warning as standard productivity hygiene. Be careful, don't burn out. But Halligan isn't talking about burnout. He's talking about the pattern that kills companies: doing too many things. And he's saying AI makes that pattern the new default.

Productivity feels like winning. Every new project you spin up feels like progress. The damage is cumulative and delayed. You can feel yourself shipping. You can't feel yourself losing focus.

By the time it shows up in the numbers, you've already built three products nobody needed.

It stays invisible because AI productivity is simultaneously an investment thesis, a career bet, and an identity marker.

You can't warn about a tool that everyone's compensation depends on celebrating. The person who says "maybe we should do less with this" is arguing against the entire room's incentive structure.

And nobody's incentivised to count.

[Lenny Rachitsky @lennysan]

Why it’s never been easier to start a company, and never been harder to scale one with @bhalligan

Brian is @Sequoia's in-house CEO coach, long-time CEO and co-founder of @HubSpot, and more than anyone I’ve ever met, a student of the job of a CEO.

We discuss:
🔸 The most common skill gap in new CEOs
🔸 Why you should build your team like the 2004 Red So
🔸 Why “When you have to eat a shit sandwich, don’t nibble.”
🔸 His LOCKS framework for evaluating founders
🔸 Much more

Listen now 👇
• YouTube: https://t.co/4RPYaZEUfs
• Spotify: https://t.co/tDOBAJv7F3
• Apple: https://t.co/nYr89fLtnG

Thank you to our wonderful sponsors for supporting the podcast:
🏆 @sentry — Code breaks, fix it faster: https://t.co/WE5wBAA74Y
🏆 @datadoghq — Now home to Eppo, the leading experimentation and feature flagging platform: https://t.co/BsR16CMQo1
🏆 @WorkOS — Modern identity platform for B2B SaaS, free up to 1 million MAUs: https://t.co/XH3bKmeEkH

[Antimemetic Insights @antimemetic_]

Dr. Tommy Wood (@DrRagnar) on The Tim Ferriss Show (@timferriss): Sleep deprivation research has a finding that nobody talks about because it contradicts the narrative.

"Processing speed slows down, but accuracy doesn't. So the quality of your work is just as good. It just might take you a little longer. And mood is affected much more than performance. So actually we don't perform less well. We just feel more grumpy about how we perform."

The sleep optimisation culture has a clean story. Bad sleep equals bad performance. Track your sleep, fix your sleep, fix your life. Wood isn't arguing that sleep doesn't matter. He's pointing at a specific gap between what the data shows and what people experience.

Short-term sleep deprivation makes you slower, not worse. And it makes you feel much worse than you actually are.

This is a measurement antimeme.

The subjective experience of being sleep-deprived, the fog, the irritability, the conviction that you're operating at 40%, is a more dramatic signal than the actual performance deficit. Your mood is lying to you about your capability. That's uncomfortable in both directions. It means the person who pulled an all-nighter and insists they're useless might be more functional than they think. It also means the entire framework of "I didn't sleep well so today is a write-off" might be mood talking, not reality.

None of this is an argument for skipping sleep. It's a narrower point. The feeling of impairment and the fact of impairment are two different things, and most people never find out because they trust the feeling.

[Tim Ferriss @tferriss]

NEW podcast episode is up!

Dr. Tommy Wood — How to Future-Proof Your Brain from Dementia

Dr. Tommy Wood (@DrRagnar) is an associate professor of pediatrics and neuroscience at the University of Washington, where his research focuses on brain health across the lifespan. This includes therapies for brain injury in newborns, prevention and treatment of adult brain trauma, and the factors that contribute to long-term cognitive function and cognitive decline. Tommy received an undergraduate degree in biochemistry from the University of Cambridge, a medical degree from the University of Oxford, and a PhD in Physiology and Neuroscience from the University of Oslo.

Alongside his academic work, Tommy is head scientist for Motorsport at Hintsa Performance, overseeing health and performance programs for multiple Formula 1 drivers. He also helped to found the British Society of Lifestyle Medicine, is head of research for the dementia prevention charity Food for the Brain, and serves as chief science officer for brain health coaching company BetterBrain. He has also trained and competed in multiple sports, coming in the top 20 in the world’s first ever fully off-road ironman triathlon, and 2nd at Washington’s Strongest Man in 2024.

Tommy is co-host of the Better Brain Fitness podcast and author of the forthcoming book The Stimulated Mind.

Please enjoy!

[Antimemetic Insights @antimemetic_]

Antimemetic Insight from @nikitabier on Out of Office (@mignano):

For years, creators built entire strategies around "the algorithm penalizes links." Newsletters, courses, workarounds. An entire cottage industry of advice on how to beat the suppression. Turns out there was nothing to beat.

"It wasn't any intentional act of like demoting links. It was just a byproduct of the UI."

The surface reading is platform conspiracy: X made a policy decision to suppress links because they push people off-platform. Every creator "knew" this. It explained their bad reach. It gave them something to game.

What actually happened was simpler. When you tapped a link, the web view opened and the engagement buttons disappeared. You read the article, closed it, went back to your timeline. The algorithm never got a signal about whether that content was any good, because the UI made engagement impossible while reading. Link posts looked like dead content to the ranking system.

This is a system antimeme. The "decision" everyone attributed to a person or a policy was an emergent property of product design. Nobody sat in a room and said "penalize links." The interface just made it structurally impossible for link posts to generate the signals the algorithm needed.

Nikita's fix was almost comically simple: keep the engagement buttons visible while people read. Link impressions went up. Time spent stayed flat. The problem that spawned years of creator strategy was a UI oversight.

Makes you wonder how many other things people build strategies around are just design accidents nobody's looked at yet.

[Michael Mignano @mignano]

The one and only @nikitabier is today's guest on Out of Office. He's @x's head of product, a @lightspeedvp venture partner, and the world's leading expert on consumer product growth.

He and I recently had a long conversation in Los Angeles where we covered the evolution of the X product, how he mastered growth by hacking and learning to "think like an adversary", and what it's like to work with Elon Musk.

Enjoy.

Chapters:

00:00 No Silver Bullet to Growth
00:18 The Evolution of X: From PayPal to Financial Network
01:25 A Tour of South Bay
01:56 Growing Up in South Bay: Internet Adventures
03:11 From Hacking to Growth Hacking
06:57 Joining X: The Journey and Challenges
09:14 Revolutionizing X: Growth and Product Development
15:44 The Role of a Product Head at X
29:13 Balancing Free Speech and Authenticity on X
34:58 The Future of X: AI, Links, and User Engagement
37:22 Solving Engagement Issues on X
40:42 Working with Elon Musk
50:13 Building Viral Apps: TBH and Gas
01:02:35 Consulting and Angel Investing
01:10:09 AI's Impact on App Development
01:12:49 Personal Anecdotes and Reflections