Basten Heutink

chief of staff @withdelphi

[Dara @daraladje]

Imagine two people using AI every day for ten years. One becomes much better at thinking. The other can barely think without it.

I’m already seeing hints of the second outcome: people producing results they can’t explain or defend. It feels like the beginning of Idiocracy.

I asked @jimkwik how to avoid becoming that second person. He’s one of the leading experts in memory and accelerated learning, and has spent three decades coaching CEOs, elite athletes, Hollywood actors, and TED speakers on how to use their minds better.

Jim shared the process he uses to keep himself thinking while using AI. It’s called BUILD:

B - Brainstorm what you already know.
U - Understand the real problem.
I - Imagine three possible solutions.
L - Leverage AI to expose your blind spots.
D - Decide what to do.

AI comes fourth. It has your thinking to challenge, and you still have to make the call.

On this week’s episode of The Library of Minds, I spoke with Limitless Daily author Jim Kwik about why learning still matters when answers are instant.

0:48 - What Exceptional Minds Do Differently
3:03 - Digital Dementia Is Real
11:05 - CrossFit for Your Mind
21:13 - Raising Kids Who Do Hard Things
33:23 - The Illusion of Learning
43:26 - The Coming Meaning Crisis
52:01 - Do This Before Asking AI
57:24 - You Can't Outperform Your Self-Image
75:38 - Ask Jim Anything, Anytime

Chat with his Delphi: https://t.co/QLcpJwL8RC

[Dara @daraladje]

We spend so much time studying how Stripe hires, how Airbnb designs, how Meta ships, how Jobs presented...

According to @joulee, the useful question probably isn't “what do the best companies do?”

It's “what are we unusually good at, and what kind of company should exist around that?”

Julie Zhuo spent 14 years at Facebook and eventually ran design for the Facebook app.

Then she started her own company and realized she couldn't copy Facebook.

Her explanation was pretty simple: “I’m not Mark.”

Companies have personalities just like people do. Their strengths and weaknesses often come from the same traits, which means copying another great company’s operating system can actually make you worse.

Really loved chatting with Julie on The Library of Minds. Join the conversation by chatting with her Delphi: https://t.co/mbo3EZziAf

0:58 - Why She Refused the CEO Title
6:48 - Delete Every Meeting but One
12:19 - The Oldest Interface Always Wins
15:40 - From Intern #1 to Facebook's Design Leader
21:08 - Writing the New Manager Bible
24:45 - How to Manage Agents

[Parth Badhwar @parth_badhwar]

Today, we're introducing @lockeai_inc, backed by @ycombinator.

Locke is going to change the way companies and governments work together.

Our first focus: building our own AI-native government affairs firm. Locke’s AI agents and in-house lobbyists help companies monitor regulations, lobby policymakers, and win government contracts faster and cheaper than ever before.

Six weeks in, we've blown past a $500k annual revenue run rate and support clients across defense, healthcare, GovTech, and more.

Government is the largest market in the world - with the worst front door. Locke will be the company to change that.

Learn more at https://t.co/mlNbEB66OL or email [email protected] to get in touch.

[Dara @daraladje]

Digital minds finally becoming mainstream

[Dara @daraladje]

Uber had $3B in the bank. Lyft had 5 months left to live.

In this week's episode of The Library of Minds, Lyft cofounder @johnzimmer tells the story of how he fought back, stood up for drivers, and refused to lose.

00:55 - My Lifetime Lyft Ban

07:00 - Uber Had $3 Billion. Lyft Had 5 Months Left.

15:08 - The Lyft Strategies Nobody Talks About

20:20 - COVID Almost Broke Him

23:47 - Tips for Founder Mental Health

27:45 - Self-Driving Cars Will Kill Uber and Lyft… Right?

36:57 - The #1 Skill Every Founder Needs

39:54 - The Real Reason He Walked Away From Lyft

[Dara @daraladje]

Was just talking about this with Lyft cofounder @johnzimmer last week

Episode drops next Thursday 👀

[Dara @daraladje]

Mercor, Lovable, Vercel, and Anthropic - all hacked in the last few weeks.

Many founders, including myself, are asking "how can we prevent this from happening?”

So I asked @EnriqueSalem, former CEO of Symantec (cybersecurity giant) and now partner at @BainCapVC.

And he said "It's not a matter of if you'll get hacked, it's only when and how often”

So don’t try to protect everything. Only focus on what actually matters.

In this week’s episode of The Library of Minds we discuss why you shouldn’t install OpenClaw on your main computer, how startups should prioritize what’s worth securing, and the nightmare scenario of agents acting with your credentials but without your judgment.

03:02 - How AI radicalized the spam problem

07:49 - How to poison an agent

09:21 - The new cybersecurity threats from AI

13:26 - The AI nightmare scenario

14:50 - Will your agent blackmail you?

16:35 - Why he trusts Perplexity and not OpenClaw

18:56 - Anthropic, Vercel, Lovable, Mercor breaches

22:41 - How to build a great security product

27:45 - Getting pushed out of Symantec

29:26 - What Enrique is hunting for in 2026

[Dara @daraladje]

If AI companies don’t start winning the narrative war the public is going to burn it all down says @braveben, ex-CRO of Flexport & partner at Saga Ventures

Not a product war. A narrative war.

AI labs are discovering cancer treatments, new molecules, and materials breakthroughs in real time.

Nobody's celebrating. People are fighting the construction of data centers that would create local tax revenue.

Ben's conclusion: The labs that are effectively becoming nation states in terms of their level of power and influence. They're going to have to make massive shows to the public or the public will burn it all down.

In this week's episode of The Library of Minds, Ben breaks down the AI narrative war, why great companies should be going public, why list construction is still the most underrated GTM skill, and why outbound as we know it might be dead.

01:15 — The story of Ryan Peterson and Flexport
05:09 — We are losing AI narrative war
10:55 — All Americans need ownership in the future
13:23 — His secret for B2B hyper growth
20:38 — The real definition of an AI company
24:28 — What it takes to change the world
27:08 — Opportunities: What Ben is looking for

[Dara @daraladje]

“If we were intelligent enough, we would never need experiments.”

@spenserskates (Cofounder of @Amplitude_HQ) thinks the current obsession with "Taste" is a trap.

In 2026, the popular narrative is that intuition is the only moat left.

Spenser believes the opposite: relying on your gut is the fastest way to get outpaced.

The winners of the AI era won't be the most “tasteful” at the start - they’ll be the ones using AI to 100x their output. High volume generates more data. More data develops better taste.

Quantity is the only real path to quality.

In this week’s episode of The Library of Minds, we discuss:

00:44 - Creating Siri before Siri

02:46 - The discovery that led Amplitude

04:10 - Why you’re not charging your customers enough

06:56 - The future of automated software development

09:01 - Apps will be personalized to each user

11:45 - Intuition is not a moat

13:21 - How Calm won the meditation industry

22:58 - 50% of startup outcomes is just not quitting

27:34 - Opportunities: What Spenser is looking for

[Dara @daraladje]

Most people will never get to ask an NBA player how they think.

Now they can.

@mosesmoody of the @warriors just launched his Digital Mind with Delphi.

Young athletes can ask about the mindset that got him to the NBA.

Fans can ask about games and decisions - anytime, anywhere

[Moses @mosesmoody]

I can’t be everywhere at once, but my Digital Mind makes it feel like I can. I partnered with @withdelphi to create an AI version of myself for my foundation. Now I can finally give back at scale.

It can talk hoops, mindset, or how to navigate whatever you're going through.

[Dara @daraladje]

At 26, @chadbyers invested $250k into Robinhood after everyone passed.

That bet turned into $400M+.

Before he wrote the check, two experts told him: “This was tried in the 2000s. It will never work.” He still wrote the check.

Later, an expert warned him off Plaid. He listened and missed out on millions.

Chad’s takeaway is not “ignore experts.”

It is this: experts explain how the system works today. They are worse at seeing when the system is about to change.

In this week’s episode of The Library of Minds, the co-founder of @SusaVentures breaks down the full Robinhood story, his unicorn filter, the new data moat, and what comes after AI.

1:45 - To make money, don’t listen to the experts

2:50 - Inventing the “data moat” thesis

5:05 - How to Find “Spiky” People

8:10 - Why 48-Hour Deal Cycles are a Mistake

10:40 - The crumbling SaaS moat

16:55 - “Oh shit, oh fuck” is the real startup journey

18:00 - How Chad raised Fund I with no track record

22:50 - When to act on your conviction

25:39 - How living w Neuralink founders shaped his diligence

27:50 - What comes after the AI wave

[Dara @daraladje]

How unicorn founder @thisisgrantlee took Gamma from “the worst idea ever” to a $2.1B company.

VCs hung up on him, friends lied to his face, and signups plateaued. Most founders would have pivoted.

Grant doubled down and went all in on the users first 30 seconds:
• The entire team focused on it for 4 month
• "Time to Value" was their key metric to drive growth

That bet took @GammaApp to 70M users who can't stop sharing it.

In this week’s episode of The Library of Minds, Grant breaks down the playbook for building a breakout AI success: staying ruthless on what matters, mastering creator-driven distribution, and hiring painfully slow.

5:10 VC: “worst idea I’ve ever heard” (then rage quit)

10:54 Betting the company on the first 30-seconds

13:05 How to run a micro-influencer program

16:14 The secret benefit to hiring painfully slow

18:12 The pains of scaling

23:29 Grant’s most challenging moment

26:45 How to launch your product

28:12 What Grant is looking for - opportunities

[Dara @daraladje]

"You can't just come in here and shit on our style"

When Ben Blumenrose (@benblumenrose) joined Facebook as its 5th designer, he tried to radically overhaul the site with high-end graphics.

Zuck shut him down instantly. The site’s "wireframe" look wasn't an accident - it was a performance feature. Minimal graphics meant the fastest load times on the internet.

Ben realized he had to understand the "Why" behind the system before he could move the needle on the "What."

He then walked away from Facebook to back a thesis the Valley dismissed: design-led companies. He built @designerfund and backed Stripe, Figma, Notion, and Linear before the world caught on.

In this week's episode of The Library of Minds:

6:40 - Moving a button cost Facebook 200M users

12:43 - Zuck: "You can't just shit on our style"

19:10 - The philosophers guiding Mark Zuckerberg

20:12 - Designing for death: Facebook's first users die

23:35 - How to be a 100x AI Designer

27:19 - "Everyone's gone rogue"

35:38 - OpenAI wont build this

[Dara @daraladje]

Most "hard tech" startups fail because they choose the wrong kind of hard.

@Farshchi (PhD ex-chip researcher & GP at @Lux_Capital) has spent 18 years investing in the most complex companies on earth (Zoox, Relativity Space, Mosaic).

He breaks "hard" into two categories:

• Scientific Risk: Fundamental discoveries that don't exist yet.
• Technical Difficulty: Engineering known science into a monopolistic product.

The mistake isn’t ambition. It’s underwriting discovery when the job is execution.

In this week's episode of The Library of Minds, we discuss how elite deep-tech investors think about risk before metrics exist, why finding a cure for cancer means nothing without distribution, and much more!

01:16 — From PhD to Founder: Choosing Hard Problems
07:11 — The "Ask Nothing" Strategy for Breaking Into VC
11:38 — Balancing Technical Difficulty vs. Scientific Risk
13:59 — Deep Tech Startups Having One Shot
18:26 — The Controversial Mosaic Investment
23:56 — Learning to Trust The Bets You've Made
27:20 — Hard Truths About Scaling from 1 → N
31:44 — Why Money Is The Ultimate Commodity

[Dara @daraladje]

Most founders accidentally train their customers never to pay them.

@MadhavanSF (the "pricing guru" of Silicon Valley - having worked with LinkedIn, Uber, and 30+ unicorns) calls it the 20/80 Pricing Trap

• 20% of your features drive 80% of the willingness to pay
• Founders give that 20% away for free to gain distribution
• You are left trying to monetize the remaining 80% of features - the ones users don't actually value.

The result? You build a charity, not a business.

In this week's episode of The Library of Minds, we discuss the science of monetization and deconstruct how to architect ‘Profitable Growth’ - the core framework from his new book, Scaling Innovation.

03:33 - Netflix vs Blockbuster: The Pricing Decision That Changed Tech
08:06 - Why Most Startups Get Pricing Wrong
11:39 - Freemium vs Paid: When Free Destroys Value
15:38 - Pricing Models Matter More Than Price
16:28 - The AI Pricing Framework: Autonomy vs Attribution
21:49 - The Biggest Pricing Mistake Ever
25:05 - Why Steve Jobs Was a Pricing Genius
26:44 - Behavioral Pricing: How Founders 10× Deals Without Changing Product
31:02 - Data vs Conviction: How Great Founders Make Pricing Decisions

[Dara @daraladje]

Culture is how people make decisions when you’re not in the room.

HubSpot’s Co-Founder @bhalligan learned that the hard way.

He built a $20B company, coined inbound marketing - and later realized culture, not code, is what truly scales.

He also drew inspiration from an unlikely source: The Grateful Dead.

They made every concert unique, gave superfans front-row seats, and even invited people to pirate their shows.

Brian built HubSpot with the same playbook.

In this week’s episode of The Library of Minds, we talk about:
• Why culture beats strategy in the long run
• Succeeding (and failing) to create new categories
• What AI means for the next generation of “inbound”

00:00 - Intro
01:47 - How inbound marketing was invented (SEO, blogging, early social)
03:38 - Naming the category: inbound vs outbound marketing
06:58 - The pivot to CRM & competing with Salesforce
09:55 - Fixing churn: the Mary / MoFu / Monetization framework
15:52 - The freemium CRM strategy & beating Salesforce from below
17:44 - Making support world-class: Apple Store hiring, culture, & training
20:01 - HubSpot’s culture code - scaling through culture, not code
27:10 - How AI is rewriting inbound marketing, SEO, and brand discovery
35:54 - The Grateful Dead: the original freemium marketers & HubSpot’s inspiration

[Basten Heutink @bastenheut]

@stanleytang @daraladje @withdelphi @DoorDash Talk to Stanley's Delphi here: https://t.co/EpWIQQ9klq

[Stanley Tang @stanleytang]

Grateful for this conversation with @daraladje & the @withdelphi team - we dove into founding @DoorDash, early scrappiness, and how much you can actually scale things that don't scale

[Dara @daraladje]

THE LIBRARY OF MINDS: EPISODE 4 - Stanley Tang (@stanleytang)

Co-founder of $DASH and one of the earliest YC founders to scale from dorm room to IPO. Builder, operator, and investor redefining how technology moves the physical world.

We discuss:
• The unexpected journey from a college project to DoorDash
• Overcoming operational challenges with creative solutions
• Embracing chaos and innovation
• The role of robotics and automation in the future of delivery
• Balancing customer experience with rapid scaling

Plus: Insights into customer love as a core value, the importance of iteration, and how @DoorDash navigates the complexities of a tech-driven operational business.

(00:00) - Intro
(01:00) - Who is Stanley Tang
(01:29) - Early childhood: computers, curiosity, and a physicist dad
(02:46) - The class project that became DoorDash
(06:38) - The first real delivery (and how it all began)
(09:08) - Product-market fit before software
(11:06) - Doing things that don’t scale - to the extreme
(15:54) - Competing against consumer behavior
(16:25) - How DoorDash built operational excellence
(20:01) - Speed vs. quality: finding the right balance
(22:53) - Robotics, drones, and the future of delivery
(26:58) - The darkest moment (and spending 40% of cash to do what’s right)
(30:34) - Chaos never ends - learning to embrace it
(31:48) - Training thousands of dashers before automation
(33:48) - Closing thoughts + talk to Stanley’s digital mind on Delphi