Weekly digest #87
Rude marketing, Why LPs should back emerging managers, Tips for better interactions
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
Software development renaissance (3 minute read)
Greg Brockman with a banger. OpenAI says AI coding tools like Codex have improved dramatically since December, now writing essentially all code for some engineers. They’re pushing internal teams to make agents the default tool (they want engineers working with an agent, not in the editor/terminal), which is a totally new way of working. He offers some tactics for how to approach it as a company.
The Anthropic hive mind (16 minute read)
Steve Yegge has been on a bit of a raid lately discussing the future of work (or rather, lack of work) in the face of AI. He spent some time with the team at Anthropic, and put this piece together exploring how they work. Describing Anthropic like a “hive mind running off vibes”, he suggests it’s the way all companies should work. Small teams in comparison to the work / scale of opportunity. No committees or traditional corporate structure, focusing on the work to be done, and not the people around it. It’s a good read.
Letters to a young person worrying about AI (9 minute read)
Addressed to young people anxious about AI taking their jobs. The argument is that you don’t need to predict the future to navigate it. History consistently shows technology displaces jobs far slower than anyone expects, because it always exists within systems of training, regulation, and supply chains that take years to switch over. Horses peaked in Finland’s lumber industry in 1950. Rickshaws spread across Asia during the Space Race. The more practical strategy is to stop predicting and instead watch what’s actually happening, then adapt. You have more time than you think.
Operational tactics
Rude marketing (3 minute read)
Tom Orbach rounds up five examples of brands being cheeky with their marketing. Standouts include Pylon showing which competitors their customers ditched, Anthropic mocking ChatGPT’s ads, and Notion subtly roasting Figma’s lacklustre product launch. The common thread is that attack marketing works best when you’re the underdog, because the bigger brand can’t clap back without looking insecure.
A new look at free-to-paid conversion (8 minute read)
Kyle Poyar surveyed 200 B2B software products to understand what good free-to-paid conversion actually looks like in 2026. The headline finding is that the old freemium vs free trial debate misses the point. Neither model consistently wins. What matters more is what you’re optimising for. Ungated experiences (letting people use the product before creating an account) drive the most adoption, while requiring a credit card upfront drives the highest conversion. The median conversion rate across all products is 8%, but almost nobody actually sits at 8%, with a 10x gap between the top and bottom performers.
Startup positioning (from HubSpot’s co-founder) (1 minute read)
Most startups are positioning themselves as “AI-first,” but that label does almost nothing for customers. It’s the same trap as “mobile-first” a decade ago. Instead of leading with the technology, frame your positioning around the value you create. Investors don’t really know what “AI-first” means either, so solving for customers solves for both audiences.
Refer and we’ll send you our favorite books as a “thank you” for spreading the word.
Private investing
Death of software. Nah. (12 minute read)
Steven Sinofsky argues that AI won’t kill software, just like the internet didn’t kill retail and streaming didn’t kill traditional media. Every major tech transition was predicted to wipe out what came before, but instead everything got bigger and more complex. We ended up with more PCs, more stores, more content. AI will follow the same pattern. We’ll need vastly more software, not less, because we’re nowhere near meeting demand for what software can do. New companies will emerge, old ones will adapt, domain expertise will matter more than ever, and the whole transition will take far longer than the “software is dead” crowd expects.
The second desert in venture capital (7 minute read)
Most attention in venture capital goes to the brutal early years, where roughly 86% of firms don’t survive from Fund I to Fund IV. But there’s a second desert waiting on the other side for those who do make it. In this phase (Funds IV through VII), the old problems are solved but new ones emerge. Your peer group thins out as former colleagues drop away. The mentors who guided you early on become less relevant. The questions shift from survival to longer-term concerns like firm longevity and succession. It’s a lonely, ambiguous middle ground where you’re no longer emerging but not yet established.
11 personal convictions why LPs should invest in emerging managers (19 minute read)
A case for why limited partners should back emerging managers rather than defaulting to established venture firms. Data from Cambridge Associates shows top-performing funds have disproportionately been emerging managers, and smaller funds can outperform larger ones by 45-60%. Because emerging managers depend on carry rather than management fees, their incentives align more tightly with LP returns. There are practical advantages too. Access is easier, the influence your cheque buys is greater, and co-investment opportunities give LPs direct exposure to promising companies they'd never reach on their own.
Career management
7 behaviors that separate juniors from seniors (engineering) (12 minute read)
The seven behaviors here are less about what you do and more about how you think. Seniors adjust their communication based on who’s in the room. They close loops instead of letting things quietly slip. They define what “done” looks like before writing any code. They think through second-order effects before committing to a solution. They know when to escalate and when to handle it themselves. They optimise for velocity over raw speed. And they anticipate problems rather than reacting to them. None of it requires extraordinary talent, it’s all learnable through deliberate practice.
Tips for better interactions (4 minute read)
A practical toolkit for how to carry yourself better in meetings and workplace interactions. The advice ranges from tactical (take notes, it forces clarity and gives you control) to psychological (never accept being labelled as “frustrated” because it reframes everything that follows against you). What ties it together is the idea that small, deliberate behavioural shifts compound into a noticeably different presence at work.
What to do when your manager doesn’t have a strategy (5 minute read)
A playbook for when your manager’s strategy feels like a shrug. The trick is realising that most leaders do have a strategy, it’s just half-formed and living in their head. Rather than pointing out the gap (which never goes well), position yourself as someone trying to align your work, interview them with the right questions, then write it all down so they can react to a draft rather than create from scratch. You end up with the clarity you need, and they look good in the process.