LLenny's Podcast
← All episodes
Andrew ‘Boz’ Bosworth (CTO)03 March 2024

Making Meta

8Frameworks
15Insights

Frameworks in this episode

Communication6 steps

Communication Is The Job

Impact happens only through artifacts and words — so map the mental model before you speak.

Read →
Self-Mastery5 steps

Identity Threat to Curiosity

Your worst behaviour comes from identity threat — turn the clench into 'fascinating, tell me more'.

Read →
Leadership5 steps

Leverage Your Leaders

Stop proving you can do it alone — make it cheap for your manager to unblock you.

Read →
Mindset5 steps

Reading Your Own Critics

Read every criticism carefully, believe none of it blindly — you know more than the critics do.

Read →
Innovation5 steps

Stated vs Revealed Preferences: Holding Conviction Through User Outrage

When users scream and keep refreshing, believe the behaviour — then debug details, not vision.

Read →
Leadership5 steps

The Eye of Sauron: Calibrating Leader Depth

Go pixel-deep only on what hinges success — and tell people exactly where their space begins.

Read →
Strategy5 steps

The Forge and the Dam

Two career positions compound: the company's hottest fire, or the critical thing nobody watches.

Read →
Peak Performance5 steps

The Six-Month Passion Test

Don't trust your identity about what you'll love — give the unfamiliar job six months, then judge.

Read →

Insights & moments

The myth-busts, hot takes, explainers, and tools worth keeping.

Myth Buster· 1

Myth Buster24:00

Meta Is More Top-Down Than People Think

The popular image of Meta as a purely bottom-up culture is, Boz says, a bit of a myth. The big ideas start with Mark and other senior leaders who have strong opinions about direction. Bottom-up creativity is real, but it operates within a top-down framework: Mark sets direction, then creates space for people to bring back several versions to shape.

  • The ideas being pursued are Mark Zuckerberg's first and foremost, though he's open to new ones
  • It's more bottom-up than most companies because of the degrees of freedom within the framework
  • Mark and other leaders hold very strong opinions about what the company should be doing
  • Bottom-up works within top-down guardrails — clarity on where you have space and where you don't

“like the ideas that we're pursuing are Mark Zuckerberg's ideas first and foremost that's not that he's not open to new ones”

Andrew Bosworth · 24:00

“make no mistake like Mark or me or Chris Cox or Javi like have very strong opinions about what we should be doing as a…”

Andrew Bosworth · 26:30
#meta#leadership#culture#decision-making

Hot Take· 3

Hot Take09:00

Startup Sacrifice Is Only Romantic in the Success Stories

Boz pushes back on startup romanticism. He gave up most of his 20s socially, worked 120-hour weeks, gained weight and drank to cope. It paid off for him, but he stresses we only hear these stories from the winners — plenty of people sacrifice just as much, or more, and it doesn't work out.

  • We disproportionately hear sacrifice stories from the successes, not the failures
  • Many people work harder or smarter and still don't get the payoff
  • He supports startup enthusiasm but insists the experience is not glamorous in the moment
  • The 'halo' around the early days is applied in retrospect, not felt at the time

“I think it's a it's a healthy thing for people to want to throw themselves into something and take that risk but it is not…”

Andrew Bosworth · 09:30
#startups#career#sacrifice
Hot Take50:00

Being Honest Isn't a Get-Out-of-Jail-Free Card

Boz says his biggest professional regrets are times he wasn't kind. As a young man he treated honesty as a license to say anything, which was wrong. Being kind isn't being nice or telling half-truths — it's delivering the same honest feedback in a way that's actually productive and helps the person rather than leaving them feeling helpless.

  • His biggest professional regrets are failures of kindness, not honesty
  • He wrongly treated 'being honest' as a get-out-of-jail-free card to say whatever he wanted
  • Being kind is not being nice, patronizing, or trading in half-truths
  • Kindness is framing true feedback so it's helpful rather than making someone feel bad and helpless

“by far my biggest professional regrets were me not being kind”

Andrew Bosworth · 50:00

“I saw being honest as like kind of a get out of jail free card like you could say whatever you wanted as long as…”

Andrew Bosworth · 50:00
#feedback#leadership#kindness#communication
Hot Take1:01:30

Why Boz Would Take the Quest 3 Over the Apple Vision Pro

After trying the Apple Vision Pro with Mark, Boz says it's genuinely great for sitting still and watching high-res movies, with beautiful UI polish and eye tracking. But the field of view is small, the displays dimmer, and motion blur distracting in mixed reality — where Meta's hand tracking is better. His frustration: most critics haven't even tried the Quest 3, which does remote desktop and multi-monitor work at a fraction of the price.

  • Vision Pro shines when stationary and watching high-res movies, with strong UI polish and eye tracking
  • Its field of view is narrow once eye relief is factored in, displays are dimmer, and MR has distracting motion blur
  • Meta's hand tracking is better and the controller adds modalities Boz values
  • His real complaint: critics judge without ever trying the Quest 3, which does remote desktop and 3-monitor workrooms

“most people haven't done that they have not tried the quest three”

Andrew Bosworth · 1:04:30

“ask yourself if You' rather have seven of those one for you and six of your best friends or one apple Vision Pro”

Andrew Bosworth · 1:04:30
#vr#apple-vision-pro#quest#hardware

Explainer· 4

Explainer23:00

Why Meta Shares Everything — and What It Costs

Boz explains Meta's open-information culture: talented people are the most valuable asset, so any time they lack information you've blocked that value and made them more likely to leave. Most companies carefully manage what information reaches people; Meta inverts that. But it isn't free — everyone has to build their own system for triaging noise and finding signal.

  • Withholding information blocks the most economically valuable thing a company has: talented people's time and attention
  • Most companies give people only the information they're 'intended to get'; Meta flips this to full access
  • Great work often comes from people having more context to act on, not from top-down mandates
  • The price is that everyone must build a personal system to triage incoming information and find signal in the noise

“most people and most companies consume all the information that's given to them but the information itself is carefully managed”

Andrew Bosworth · 25:00
#meta#culture#transparency#information
Explainer30:30

The 'Eye of Sauron': When Leaders Should Go Deep

Boz shares Meta's internal metaphor for Mark's attention: when Mark decides something is the most important thing, no detail is too small — he'll critique strategy and a single wrong pixel in the same review. The lesson on how deep a leader should get is 'it depends on the weeds' — some weeds decide success or failure, most don't. And being under (or outside) that intense gaze both carry a cost.

  • When Mark deems your work the most important thing, there's no detail too small for him to notice
  • Founders should distinguish the weeds that are the hinge of success from the ones that merely have to get done
  • Boz and Chris Cox go 'high to low' on things that matter a ton, and let others iterate into rough shape
  • You suffer in the intensity of the gaze, but you also suffer in the shadow of its absence — there's no perfect place to be

“when Mark has determined that the thing that you're working on is the most important thing there is no detail too small for him to…”

Andrew Bosworth · 31:00

“you suffer in the intensity of the Gaze of Sauron you also suffer in the shadow of its absence”

Andrew Bosworth · 33:00
#leadership#founders#delegation#meta
Explainer51:30

Communication Is the Job — You Cannot Not Communicate

Boz unpacks his classic essay 'Communication Is the Job.' Impact on the world happens exclusively through artifacts and verbalizations that affect other humans — a great idea unshared means nothing. He ties it to extreme ownership: when something goes wrong, ask what you could have communicated better. And silence itself is a message — you cannot not communicate.

  • Impact is achieved exclusively through communication artifacts that affect other humans
  • A great idea you didn't break through with is on you, not the audience
  • Org charts are communication devices, not physical strings — tools for organizing who goes to whom
  • Silence and the absence of check-ins carry meaning; you are always communicating something

“if you want to have an impact on the world around you it is exclusively done through the the the creation of artifacts”

Andrew Bosworth · 52:00

“you cannot not communicate you are always communicating something with your face with your clothes with your body”

Andrew Bosworth · 54:30
#communication#leadership#ownership#management
Explainer1:11:00

You Know More Than the Critics (But Read Them Anyway)

On Meta's stock crash and recovery, Boz shares how they insulate from external opinion. Lou Holtz's line — you're never as good or as bad as they say — guides them. Internally he reminds his team they know more than analysts, media, or Twitter do. But you still read all the criticism carefully, watching for the parts you're inclined to resist that may be true, aware of Gell-Mann Amnesia.

  • Meta had the largest single-day stock drop in history, then the largest single-day gain ~18 months later
  • Lou Holtz: you're never as good as they say when winning, never as bad as they say when losing
  • You know more about your company's real value than critics, analysts, or media — but don't ignore them
  • Gell-Mann Amnesia: read everything as a point of view, watching for true critiques you'd rather resist

“you're never as good as they say you are when you're winning and they're never as bad as they say you are when you're losing”

Andrew Bosworth · 1:11:30

“you know more than the critics do you know more than the analysts in the marketplace do”

Andrew Bosworth · 1:11:30
#leadership#criticism#meta#resilience

Story· 5

Story10:00

The Unglamorous Early Facebook Grind Nobody Sees

Boz describes the reality of joining Facebook as one of the first ~16 engineers in 2006: no experts to ask, no ops team, building your own servers by driving to Fry's Electronics. He built an anti-spam/anti-scraping tool himself and then had to wake every four hours for about two years to fend off attackers. The point: the romantic startup story is real but the sacrifice is invisible in the retelling.

  • He and five others joined the same day, so everyone claims to be the '10th engineer' — really somewhere between 10 and 16
  • There was no expert to consult and no 24/7 ops support, so problems fell on whoever noticed them
  • When servers ran out of capacity, you drove to Fry's, bought components, and built a new server yourself
  • He built the anti-spam defense unprompted and woke every 4 hours for ~2 years to check for attacks

“I didn't sleep for more than four hours at a time I had to wake up every four hours and check the report and see…”

Andrew Bosworth · 10:30

“they don't tell you about that stuff in the movies”

Andrew Bosworth · 10:30
#startups#facebook#engineering#early-days
Story33:30

Newsfeed: Conviction and Stated vs. Revealed Preferences

Boz recounts building the first algorithmic news feed, which users initially hated. The key signal was cognitive dissonance: everyone said they hated it while immediately doubling their usage — stated preferences vs. revealed preferences. The lesson isn't 'ignore customers'; it's having conviction in a vision users will grow into, while diagnosing whether the thing itself is wrong or just the details.

  • Users were outraged at Newsfeed at the exact moment they doubled their usage of it
  • The tell was the gap between stated preferences and revealed preferences in the economic sense
  • You choose your customers as much as they choose you — early cohorts can pull you toward the wrong market
  • The art is diagnosing whether the thing itself is wrong versus whether only the details were wrong

“everyone was outraged at the same time as they immediately doubled their usage of the product”

Andrew Bosworth · 34:30

“there's cognitive dissonance here uh between what the stated preferences and what the revealed preferences are”

Andrew Bosworth · 35:00
#product#facebook#newsfeed#conviction
Story45:00

How Mark Zuckerberg Takes Feedback: Recompile Overnight

Boz describes the uncanny experience of giving Mark feedback: Mark listens genuinely, most often tells you you're wrong and why — and then over the following week or two, you quietly see him shift. Boz's analogy is that Mark absorbs enormous information daily and 'recompiles the whole world' overnight. It's a live example of strong opinions loosely held, and it applies to Mark's own presence and delivery, not just product.

  • Mark truly listens, then usually tells you why you're wrong in the moment
  • Over the next week or two you see the shifts happen anyway — feedback works on a long view
  • Boz's metaphor: Mark recompiles the whole world overnight with the day's information
  • Mark also pressure-tests feedback by asking others their view, without attribution, across his next meetings

“he'll listen truly he'll most often tell you that you're wrong why you're wrong”

Andrew Bosworth · 45:30

“at night he like recompiles the whole world with all that information and comes back”

Andrew Bosworth · 46:00
#feedback#leadership#mark-zuckerberg#meta
Story1:21:30

Identity Threat: The RPC Argument Boz Still Regrets

Boz tells the story of a heated argument with his close friend and brilliant engineer Dave Fetterman over Thrift's encoding — Boz wanted efficient binary encoding to save scarce memory bandwidth; Dave argued for scrutable plain-English logs. Boz got red-faced and furious in front of half the company's engineers. Dave was right within a year, and the real lesson was about identity threat: his identity was wrapped up in being right.

  • The fight was over Thrift RPC encoding: binary efficiency vs. human-readable, scrutable logs
  • Dave Fetterman was right — within a year memory bandwidth stopped mattering versus log readability
  • Your worst behavior emerges when you feel a core part of your self-conception is under threat
  • It became a step in learning that the goal is to be open-minded and curious, not to be right

“your worst behavior is always going to come out when you think you are under identity threat”

Andrew Bosworth · 1:24:30

“within a year the memory M definitely didn't matter”

Andrew Bosworth · 1:24:00
#engineering#identity#self-awareness#facebook
Story1:27:00

Ami Vora's Superpower: Meet Disagreement With Curiosity

Boz's greatest lesson came from watching Ami Vora. Faced with a profound disagreement — even something she thought was crazy wrong — she'd respond, genuinely, 'fascinating, you have to tell me more.' Her heartfelt curiosity tore down walls between viewpoints and made people open-minded, and she'd sometimes change her own mind. Boz now tries to model it: when he feels an internal clench, he reaches for curiosity.

  • Ami Vora met even 'crazy wrong' disagreement with genuine curiosity, not defensiveness
  • Her sincere 'tell me more' disarmed people and made them lean in and open up
  • She was usually right, but she'd also change her own mind when the other view held
  • Boz now converts a strong internal 'clench' into curiosity — a kinder form of 'yes, and'

“she would say fascinating you have to tell me more more about why you think that”

Andrew Bosworth · 1:27:30

“when I have a strong strong internal clench I try to embrace curiosity”

Andrew Bosworth · 1:28:30
#communication#curiosity#leadership#self-improvement

Takeaway· 2

Takeaway13:30

Ask for Help More: Directly Leverage Your Leaders

Boz's most-frequently-given advice is that people should more directly leverage their managers, mentors, and advisors. We resist asking for help because we equate it with admitting failure, but your job is not to do it yourself — it's to get the thing done well. A manager can bulldoze blockers and correct misunderstandings that otherwise waste months.

  • People avoid asking for help to prove they can do it themselves, treating a request as an admission of defeat
  • Your actual job is to get the thing done right, and the tools to do that often live with your manager
  • A quick check-in lets a leader clear paths and fix miscommunication before months are wasted
  • Your leader is rooting for you to succeed — your success makes their life easier

“more often than not your job is not to do it yourself your job is to get it done”

Andrew Bosworth · 14:00

“no one wants you to succeed more than the person you are auditioning for”

Andrew Bosworth · 16:30
#management#career#leadership#communication
Takeaway39:00

Optimize for Learning: Give a New Role Six Months

Boz had an unusual career arc — changing jobs roughly every six months across Integrity, Newsfeed, infra, bootcamp, messaging, and groups — before going 'vertical' with ads. His advice: optimize aggressively for learning, and don't assume you won't be excited by something new. Peers who stayed in one track got promoted faster early, but some became domain specialists who exhausted their field and got stuck mid-career.

  • He moved roles whenever he stopped learning or got bored, building a wide variety of skills
  • Karate Kid metaphor: painting fences and waxing cars looked slow, but at the end he 'knew karate'
  • Don't pre-judge that you won't be passionate about a domain — give it six months to find out
  • Early career, optimize for learning like compound interest: unimpressive for ~10 years, then it compounds

“I just was willing to learn aggressively I would move because I wasn't learning enough I was bored”

Andrew Bosworth · 40:30
#career#learning#growth#meta