2025-12-11 11:24:03
So much to read but only 504 hours left in the year… gulp
How I Screwed Up Sales Hiring [Pete Warden/MoonshineAI] – Stop me if you heard this one before – the incredible technologist CEO/founder didn’t get sales motion/hiring right first time he tried. It’s a cliche but still very much real. We’re friends of, and small investors in, Pete’s company – I read his post-mortem eagerly because he’s a gem of a human, super smart, and dogged in getting MoonshineAI to where he wants it to be (moonshine is a speechAI model that runs locally).
Poems Can Trick AI Into Helping You Make a Nuclear Weapon [Matthew Gault/Wired] – Oh this is fantastic. Writing in poetic structure can make it easier to jailbreak some LLMs into giving forbidden answers. Adversarial poetry!
Lessons for New Leaders [Molly Graham/Glue Club] – So many good bits here. “Every time I’ve stepped into a new leadership role, I’ve made the same mistake: I’ve gone too slowly.” A bunch of advice and learnings for your first 30-60 days in a new org/role.
Startup Operating Principles [Sam Gerstenzang/Sam’s Newsletter] – Sam currently runs Boulton & Watt, described as the world’s slowest startup incubator!!! He previously did a bunch of stuff at Google, Stripe and another startup he cofounded. This SOP post is a follow-up to one he wrote about Stripe’s Operating Principles.
How to Fix a Typewriter and Your Life [Kurt Streeter/New York Times] – Prepare to cry. This is beautiful. When serendipity and intentionality collide, what do you do? Here our protagonist makes a decision which changes two lives. Please just read it.
Ciao for now!
I’m grooving on Blackbird, this new local restaurant/bar app which is some combination of loyalty program, crypto payments, etc – it’s curious and I’m trying to make sense of it all. You get free somethings for trying it too. Here you go:
Download Blackbird and earn 2,000 $FLY when you check in for the first time. That’s $20.00 toward any meal!
Use my code bb-o9edk2 https:/
2025-12-09 03:45:01
Prioritize ownership over valuation on any given deal, but price drives your portfolio construction. This was basically how we managed Homebrew during its first decade, when we were investing out of three institutional funds, all sub $100m. The guiding principle changed in 2022 given our evergreen personal capital model (as I detailed in Why I No Longer Care About Startup Valuation When I Invest (Except for These Four Reasons)), but it served us well when we were deploying in a more typical lifecycle.
Our strategy for Funds I, II, and III were basically to lead/co-lead seed rounds with a 10-15% ownership target. Our third fund averaged 11.6% but it was really more bimodal – a bunch of 8-12% combined with a roughly equal number of 15% – across 31 startups. The concentration of dollars (expressed in ownership) was primarily about being able to also focus our time working with the portfolio and keeping our funds modestly sized where $1b outcomes could make a real difference to returns.
So our thinking was basically something like this in Fund III [2018-2021, $90m, 31 companies]:
Did we ever say ‘no’ to an investment opportunity based on pricing? Sure, but it was more often a function of the overall round size becoming too big for us to write a lead check, rather than the difference between, say, $1.5m and $1.8m to get 15% ownership.
Much like the Mike Tyson quote “everyone has a plan until they get punched in the mouth,” every firm has a model that then gets hit by reality (aka The Market). You can decide what you put in a spreadsheet cell a year ago is gospel, or you can use it as a guidebook, and behave more dynamically given what occurs over your deployment period. But if you make every investment decision without some framework to understand what it means for the fund itself, you might find the whole becomes less than the sum of its parts.
Footnote: One of the fun parts in backing new venture firms via Screendoor, is talking with them about portfolio construction and seeing the different strategies/playbooks they implement to win.
2025-12-01 07:45:50
You might be eating Thanksgiving leftovers, but these URLs are fresh off the vine. Enjoy!
Why the New York Bodega is Here to Stay [Anna Kodé/New York Times] – GoPuff will not replace us! Doordash will not replace us! Bodegas provide community, geo specific goods and services, and a pathway for immigrants to work their way into the local economy. They are forever good. And besides, if they disappeared, where would all the Bodega Cats go?
The Production Capital Mosaic: Financing the New Industrial Deployment Age [Brett Bivens/Venture Desktop] – Brett reviews many of the emerging ways we think of providing capital to businesses and creates an important division: is the capital vehicle shaped like a company or a financing entity? Worth reading in full to really understand why he thinks this is a valuable distinction.
It’s the End of the (ARR) World and I Feel Fine [Brett Queener/Tales from the Bonfire] – Brett starts with a banger [“ARR was great for the SaaS age, but it may be as relevant as a Commodore 64 for the AI software era.”] and goes from there. He touches on impact for founders and execs; venture capitalists; and private equity models. If the AI world is about hiring, not buying, software, what else changes?
The Monks in the Casino: A brief theory of young men, “the loneliness crisis,” and life in the 21st century [Derek Thompson] – This is soooo good and starts threading together concerns Derek has about what economic degen as a way of life means for society, especially men.
This shift has moral as well as economic consequences. When a society pushes its citizens to take only financial risks, it hollows out the virtues that once made collective life possible: trust, curiosity, generosity, forgiveness. If you want two people who disagree to actually talk to each other, you build them a space to talk. If you want them to hate each other, you give them a phone.
Do Something Important, Quickly [Yoni Rechtman/99% Desirable] – “All VCs care about is the asymmetry of being right/making money without the risk of losing face.” At least he front stabs us instead of backstabs. He’s not wrong. And it’s a problem that is making our industry boring.
Why every country needs to master the Electric Tech Stack [Noah Smith/Noahpinion] – Most frameworks become overloaded as the creator tries to make everything fit their worldview. TBD whether the ‘Electric Tech Stack’ falls victim to this gravitational pull, but for now I like how it pulls together a set of industrial, economic, and technical capabilities that might define the coming decades. Batteries, Electric Motors and Power Electronics.
Enjoy some reading!
2025-11-25 03:09:07
Slinging URLs for your Thanksgiving* Week reading (*where applicable 
)
Strong Capitalism: Playing Entrepreneur on Hard Mode [Philip Rosedale, Entrepreneur] – Philip continues to pull forward the Kapor/LongNow/StewartBrand/HomebrewComputerClub ethos of early SV, here encouraging a host of values that have inspired me. It’s comedy/tragedy that he has to brand this “Hard Mode” but I get where’s he coming from – the shortcuts, shirking of responsibility, getting yours at others’ expense *does* feel like lazy, soft, uninspired building. Philip names a bunch of Operating Principles for Hard Mode – you should read the whole list. But let share his full opening paragraph to give you a sense of where he’s going.
The strong capitalist recognizes that free markets are the best way to create opportunity, innovation, and economic prosperity, while simultaneously acknowledging that in the absence of any other operating principles, capitalism leads to non-meritocratic wealth inequality, monopoly, resource depletion, and extraction. The strong capitalist demonstrates and upholds a set of additional behaviors which maximize the health, prosperity, and longevity of the free markets, even when those behaviors make their own success more difficult – AKA – “hard mode”. The strong capitalist is an honorable competitor, a source of inspiration to others, and deserving of their returns.
I Worked All Over Silicon Valley. This is How it Lost Its Spine [Aaron Zamost/New York Times] –
“For tech companies, courage doesn’t scale.” Gutshot and he’s not wrong. I had the pleasure of working with Aaron at YouTube (he later went on to executive roles at Square and others) – he’s not a cynic by default – Aaron has put his sweat and reputation into these companies. He believes in the empowering potential of technology. But maybe no longer is as confident in the technologists.
For years, Silicon Valley symbolized progress. Its retreat from its core values leaves no clear heir — no other industry fights for the future in the same way. When tech is the villain instead of the hero, the future feels leaderless. And a country that stops believing its innovators can make the world better stops believing in much else, too.
don’t take the bait. vice signaling is eating silicon valley. [Jasmine Sun/@jasmi.news] – Ok so this is the *third* link in a row that notices the gaping fissures in our community. But let me tell you why I’m encouraged, rather than depressed – these are all practitioners, not just pundits. These are people who have done – and continue to do – real work of value. These are people looking to help remodel the house, not just criticize the architecture.
I also worry that Silicon Valley now punishes outward earnestness or virtue; young technologists have expressed fears of appearing soft—or worse—woke. I don’t want these vice-signalers to represent the industry, giving credence to already budding distrust of tech. If the bubble pops, or if an SBF-style scandal erupts, or if we get evidence mapping social crises directly to AI, the public will not be kind to those who got rich bragging on the way.
New Stack, Same Game. Why the AI Native Era Feels Different Yet Familiar [Ashley Smith/Vermilion Cliffs Ventures] – Author credibility check…. ok, Ashley had marketing leadership roles at Twilio, Parse/Meta, Gitlab, Github. Verdict: Credible. Now she’s full-time on her own venture firm, Vermilion Cliffs. Here she challenges one prevailing take: “There is a common assumption that someone like an OpenAI will simply build everything. The same assumptions appeared about AWS, GitHub, and Atlassian in their key eras. But every time the platform layer stabilizes, specialization accelerates.”
Importantly she’s not just writing her beliefs, she’s funding them.
How Screendoor became a key signal for emerging VC talent [Allie Garfinkle/Fortune – $] – Screendoor backs founders who are building venture firms. It’s often like backing seed stage companies – you need an eye for talent, stay close to the ground, and be willing to commit before others believe. Here Allie takes note that our portfolio is quite remarkable; that we’re often the first institutional LP to sign on to fund; and that when we invest, other LPs fast follow. Lots of work to do but proud of what Lisa and team are building.
Enjoy!
2025-11-09 02:05:37
The most powerful force in the world is compound interest. The second most powerful is the compounding of friendship – the chance to know people deeply across very long periods of time. A note landed in my inbox this summer from a product manager who was leaving Google after more than 15 cumulative years. We had the good fortune of overlapping at YouTube – in fact, I brought him over from Google. In the recent update email he shared a historical document – an ‘ask’ he had during that recruitment period long ago – and mentioned how much it stuck with him.
Although there were certainly things I would do differently as a leader given my own personal growth since those days, I did always take talent recruitment and development very seriously. Today when great young team members have so many options of where to work, one advantage you have in attracting and retaining them is really listening to/soliciting/helping them develop career plans. And then living up to the commitments you make to them, even if it’s aimed at helping them leave when the time is right.
Sharing a snippet from the email here with his permission
2025-10-19 22:51:50
Encountered this ‘use scale’ at a high school for how students are permitted (even sometimes encouraged) to use Artificial Intelligence tools in their work. Struck me as very smart when applied with thought on a per subject, per assignment basis. Wonder what the teacher-facing guidelines are too!
