Weekly digest #91
Bits in, bits out, Principles of seed investing, I put every person in my Wharton MBA class into a spreadsheet
We (Matt and Shray) read hundreds of articles on company building, venture investing, and self-management and curate the best ones into a weekly digest to help founders and operators stay on the top of their game.
Better thinking
Bits in, bits out (11 minute read)
Erik Hoel argues that six years into the LLM era, the best measure of AI’s impact is what happened to writing, the domain where these models should shine most. The result is a flood of mediocre content, modest efficiency gains for already-skilled people, and no meaningful improvement at the top. AI is a tool, not a new source of intelligence. His case against the singularity hype is that the same promises were made about writing, and writing got worse on average. And so he expects the same pattern to repeat across every discipline AI touches. I don’t know, this feels like a skills issue to us.
White collar goes blue (14 minute read)
For decades, white-collar work carried a quiet promise that your opinion matters, your autonomy is protected, your personhood is part of the deal. AI is making intelligence abundant, and abundant things don’t command premiums. The professional laptop class is facing its biggest identity crisis since industrialisation, and most people haven’t clocked it yet. What a start to the week.
The future, according to Sequoia (32 minute watch)
Alfred Lin, Partner & Co-Steward at Sequoia Capital, talks with Molly O’Shea (founder of Sourcery VC) about the future. Wide ranging talk, which of course gets into AI. Perhaps the most intriguing thing, and somewhat counterintuitive, is how he argues that the next generation of companies (not the existing ones) could be even bigger than those before. He even stipulates we’ll see our first 10 trillion dollar company from this cohort.
Operational tactics
Every PM should be building skills (10 minute read)
Aman Khan, Head of Product at Arize AI, spent the last few months using Claude Code’s skill and plugin system to turn his coding assistant into something that actually remembers his codebase, workflows, and preferences. What he also realized is that skills are a peek into one’s mind as a builder. A well-written skill encodes how one thinks about a problem, what conventions they follow, and what mistakes they’ve learned to avoid. Though documentation exists, practical guidance is scattered. Read this post if you’re interested in learning more about skills, where to find good ones, how to make Claude (Code) invoke them reliably, and what Aman’s setup looks like.
Services, the new software (10 minute read)
Neat one from Julien Bek at Sequoia. He argues that AI companies that sell outcomes will beat those that sell tools. When you sell the work rather than the software, every model improvement makes you faster and cheaper rather than obsolete. The biggest opportunity is replacing outsourced services (accounting, legal, insurance, IT) where budgets are six times larger than software spend and buyers are already purchasing outcomes, not tools.
Revenue addiction kills companies (15 minute read)
Companies that optimise purely for short-term revenue are killing their ability to take the long-term bets that AI disruption demands. They squeeze customers, crush innovation, and poison culture, often years before the damage shows up in the numbers. Revenue should be the outcome of building something people love, not the thing you’re building toward.
Refer and we’ll send you our favorite books as a “thank you” for spreading the word.
Private investing
Founders, your software company is not going to IPO (4 minute read)
Most software startups won't IPO. The bar keeps rising (you'll need $500M+ ARR by 2030), private capital is abundant, and public markets are increasingly brutal to lower-growth companies. When your private valuation overshoots what public markets will pay, you're stuck in limbo for years, watching good people leave because they can't see upside in their equity, while trying to restructure and grow at the same time.
Principles of seed investing (2 minute read)
“Most of what is known about seed investing is borrowed from venture capital. That is often a mistake.” Well, so this post argues. They’ve outlined 10 principles to follow if you’re seed investing. Most we strongly agree with.
Career management
I put every person in my Wharton MBA class into a spreadsheet (8 minute read)
If you ran into a classmate from university at a coffee shop in a five years, what would you do? Would you hug them? Would you do a weird half-wave where you’re not sure the other person remembers your name? Vayum, a Wharton MBA alum and member of technical staff at Weco AI, thought about these questions a lot. MBA students are frequently told that their degree is about the network, relationships, and people. These are the assets that appreciate and compound over time. So he put it to the test by categorizing 805 people on a spreadsheet from ‘core circle’ to ‘never met’. What he realized was that how he’d been spending his time didn’t map to what he’d say matters to him if someone asked. For all it’s clinical absurdity, the spreadsheet made time and energy allocation visible. What you do with the info is entirely up to you.
Become a tinkerer (2 minute read)
Fast-growing startups no longer need just a marketer or engineer anymore. They want tinkerers, operators who cross functional boundaries and own outcomes end-to-end. The marketer’s in Claude figuring out what the last deploy changed, building the landing page herself, writing ad copy, and launching the campaign before day’s end. Top early-stage startups have always operated this way. The difference now is that AI lets you maintain that archetype scrappiness even as headcount scales. Companies that don’t revise their org charts and empower their peers to cross functional lanes will bleed their best people to ones that already have. Picking a lane is officially dead.
50 lessons from 50 years lived (5 minute read)
Deb Liu, former Ancestry CEO and Facebook VP, is not only a well-respected Silicon Valley executive but also a prolific blogger who’s released 250 issues over the course of 5 years. Her Substack, ‘Perspective’, has become a personal guide on how to achieve more and live more through lessons from tech, leadership, and parenthood. Across all her posts, she details lessons in categories like career foundations, incentives, and mentors. Here are the 50 lessons that mattered the most. You might want to put the kettle on for this catalogue.