The Learning Corner by Precursor

This week on The Learning Corner, an observation from investor Matt Turck that startups today are either AI-native rocketships or left for dead opens up a real conversation about what that binary actually means for founders building outside the hype cycle. Michelle Volz's essay "To Be or To Do" uses military legend John Boyd's framework to argue that Silicon Valley has quietly shifted from rewarding builders to rewarding the performance of building, and why discipline is the actual path to the biggest outcomes. We finally close with Vinnie Lauria, Founding Partner at Golden Gate Ventures, who breaks down why a VC who already passed on your company is one of the worst people to introduce you to another investor and why credible signal matters far more than access in fundraising.

This week on The Learning Corner, Mia and Charles dig into Gil Dibner's argument that Series A investors have fundamentally abandoned their core function, opting instead to invest way too late, way too early, or simply accumulating options without conviction. They then unpack Sequoia's twenty-year hype analysis that shows the most valuable companies are almost never founded in the peak hype year of their category, and what today's signals might tell us about the next big wave. The episode closes with Maple VC's Talent Arbitrage framework and why the Builder, the nonlinear, hardest-to-categorize founder, is the most important and most overlooked signal in early stage investing.

This week on The Learning Corner, we dig into three pieces that all orbit the same uncomfortable question: who is actually accountable when things go wrong? Angela McNeal's "The Architecture of Blame" argues that AI does not invent accountability gaps, it industrializes them at machine speed across every workflow at once. New academic research reveals that VC-backed startups are statistically more likely to commit fraud, and that investors are not passive victims but active co-creators of the conditions that make it happen. We close with a Wall Street Journal investigation into how the AI talent war is turning founders into acquihire targets, with VCs now explicitly asking in diligence: will you quit on me?

This week Mia and Charles are joined by Precursor intern Aleena to dig into three good reads. First up, Ashley Smith's piece on the pressures young founders face today, including predatory incubators that trade small checks for outsized equity stakes. Then, YC's Fall 2026 Requests for Startups, featuring the first ever entry from a sitting US Secretary of the Army. They close with Frontier, a piece unpacking why AI labs stopped calling their models foundation models and started calling them frontier models, and what that shift reveals about how durable their moats really are.

This week on The Learning Corner, Mia and Charles dig into Nick Chirls' case for patience as an underrated venture strategy, using the owners versus tenants framing to explain why short term thinking is eating the industry. They then explore David Deming's argument that humans still hold a real edge over AI in social learning, especially in reading people and building trust from just a handful of interactions. The episode closes with George Sivulka's provocative piece comparing AI agents to a workforce, drawing parallels between tokenmaxxing and headcount bloat, and looping and meetings about meetings. Tune in for a conversation on long term thinking, the future of human judgment, and how to actually manage an AI agent workforce well.

This week Mia and Charles dig into whether using AI tools means quietly giving away your most valuable knowledge, what Sierra learned building a single company wide AI agent instead of one per department, and why data, not tokens, might be the next trillion dollar category in AI. They also touch on New York's new one year moratorium on large data center construction. Tune in for a debate filled conversation about what it really means to adopt AI at work today.

Is AI venture living through its biggest bubble yet, or its biggest winner-take-all cycle ever? Charles and Mia break down Samir Kaji's read on the duality driving valuations to record highs. They also dig into a Stanford field experiment proving cold pitches to investors work far more often than founders think, and challenge Tomasz Tunguz's take on why most startups don't actually have a real moat at founding. A fast, sharp look at three ideas shaping venture right now.

This week we dig into a piece from Halle Tecco exploring how healthcare VCs are splitting into two camps on ownership discipline as AI drives valuations to decade highs. We also break down Sam Lessin's argument that the SpaceX IPO signals the end of the DCF as the only globally scalable story of value, and what it means that minority belief systems can now pool conviction at trillion-dollar scale. We close with Sarah Guo's "The Untrainable," a sharp framework for understanding where real moats exist in the AI era and why the defensible work is exactly the stuff that can't be benchmarked. Tune in for a packed episode touching on venture math, narrative-driven markets, and what it actually takes to build something durable when the models keep getting smarter.

This week we featured Tony Fadell's interview with Lenny Rachitsky, covering how great products are built with pain at the center and why fast software is the new fast fashion. We also dig into Hunter Walk's practical framework for how early-stage VCs should communicate SAFE note markups to their LPs. And we close out with Reuters' reporting on Kirkland and Ellis committing $500 million to build a fully proprietary AI platform, and what that signals about where the legal industry is heading.

This week on The Learning Corner, economist Alex Imas makes a counterintuitive case that AI will not eliminate human labor but instead relocate scarcity toward a "relational sector" of nurses, teachers, craftspeople, and care workers where human presence is the product itself. Lisa Kostova shares one of the most honest accounts of vibe coding gone wrong, arriving at a major conference with 40 ready buyers and a product too broken to sell. We close with Boris Cherny, creator of Claude Code at Anthropic, who argues the title of software engineer is dissolving but predicts 100 times more people will be writing code in the near future. Three reads, one big question: what does human work actually look like in an AI economy?

This week on The Learning Corner, we dig into Rachel Karten's viral piece on AI-obsessed leadership and what happens when organizations adopt AI without any real strategy. We then discuss Karan Dhir's framework for career bets that actually compound versus the ones that just look like progress, including a debate on whether the AI era has quietly rehabilitated the generalist. We close with Charlie Warzel's Atlantic piece on AI malaise and whether the anxiety around AI is actually a geography problem more than a technology one.

This week on The Learning Corner, Charles and Mia discuss a good read explaining why the quiet builder playbook is no longer enough and why founders must now own their narrative to win. They dive into a provocative take on the job market arguing the signal collapsed, not the opportunities, and what that means for the next generation of candidates. They also explore the rise of the High-Impact Individual Contributor and how AI is reshaping career paths at every level. Plus a quick flash on OpenAI's reported plans to file for IPO as early as September at a valuation of up to one trillion dollars.

This week on The Learning Corner, we open with an argument on why traditional networking culture is mostly negative selection and why broadcasting your work publicly is the stronger play. We then dig into why AI is making the "what" of your work more important than the "how," and which skills actually become load-bearing in that world. We close on a sobering piece about AI psychosis spreading through executive suites, the sycophancy loop baked into AI tools, and what it means when the feeling of running a massive organization is completely disconnected from what is actually shipping.

This week on The Learning Corner, Charles and Mia dig into who actually owns the AI agents you build at work and whether you can take them with you when you leave. They unpack what happens to a founder's standing inside a VC firm when their partner walks out the door, and why that moment is really the start of a new fundraise. They close with a sharp framework for separating companies that are truly AI-native from those just using better autocomplete. Three great reads, one tight conversation.

This week on The Learning Corner, Charles and Mia dig into Lucas Vaz's viral thread arguing that the era of easy, low-priced, diversified venture investing is over and that seed fund math is fundamentally broken. They also break down Garry Tan and YC's official guidance on why founders need to stop conflating LOIs, GMV, and ARR before it costs them investor trust. The episode closes on China's decision to block Meta's $2 billion acquisition of Manus AI and what it signals about the closing window for cross-border AI deals.

This week on The Learning Corner, Manny Medina, co-founder of Outreach, shares his biggest regret as a founder and why he believes killing your competition is the actual job description of a VC-backed founder. Samir Kaji breaks down his fundamental truths in venture capital today, touching on the widening barbell between large and small firms, ARR reporting concerns, and honest uncertainty around AI. We close with the bombshell SpaceX and Cursor partnership announcement, unpacking what a $60 billion acquisition option and a million H100 equivalent supercomputer really signals about where the AI arms race is headed.

This week we are diving into whether domain expertise is becoming more important in Vertical AI, not less, breaking down new data from The Euclid Ventures team that challenges the narrative that youth and technical ability are all you need to build a winning company. We then get into Micah Rosenbloom's breakdown of seed valuation inflation and what it means for early stage return math in today's market. We close on Kyla Scanlon's piece on AI and academic cheating, and what it really signals about the generation of workers and founders entering the workforce.

This week on The Learning Corner, Mia and Charles dig into a piece by Ashley Smith of Vermilion Cliffs Ventures on why seed round names have become essentially meaningless, with YC Demo Day valuations hitting $40M post for companies with eight weeks of runway. They then unpack Auren Hoffman's sharp and practical breakdown of how to actually ask for advice without wasting people's time, and why closing the loop is the most underrated relationship builder in your network. The episode closes on a major structural shift in venture: family offices and private wealth firms cutting out VC middlemen and going directly into startups, as covered in a recent TechCrunch Equity episode. Tune in for a conversation spanning venture pricing, shifting capital dynamics, and the soft skills that still matter in a world full of AI answers.

This week on The Learning Corner, Ashley Mayer makes the case that tech's biggest players have forfeited the narrative and it is now up to startup founders to give people a reason to root for the future. Pavel Prata at Murph Capital argues that the value-add model that transformed venture is now playing out one layer up at the LP level, and most allocators haven't caught up yet. We also dig into Jenny Fielding's take on the return of agile funding and whether stacking small SAFEs in an AI-driven world is a founder-friendly evolution or a short-term optimization with long-term consequences. All that and more on this week's episode.

This week we cover the explosive whistleblower exposé on Delve, a YC-backed compliance startup accused of fabricating audit evidence and leaving hundreds of clients unknowingly exposed to legal liability. We then dig into Jerry Neumann's "We Have Learned Nothing," a data-backed argument that decades of startup frameworks have produced no measurable improvement in whether companies actually succeed. We close with the Wall Street Journal's reporting on OpenAI's major strategic pivot away from its "do everything" approach to refocus on coding and enterprise, and what that signals about competition with Anthropic.

This week on The Learning Corner, Om Malik makes the case that reputation and perceived authority now convert faster than financial capital, and that the tech ecosystem has fully internalized that game. We then dig into a provocative LinkedIn post from CRV's Reid Christian, where CEOs are starting to build token output per month into formal performance reviews as a core KPI. We close with a thoughtful piece from Deciens on intellectual lineage and why choosing whose shoulders you stand on is one of the highest leverage decisions a founder or investor can make. Three reads, one tight conversation.

This week on The Learning Corner, Micah Rosenbloom from Founder Collective challenges the viral narrative that VC returns suck and asks whether we are all just spinning data to fit our own models. Jeff Weinstein draws a striking parallel between today's AI funding frenzy and the Nifty Fifty stocks of the 1970s, arguing that being right about a company has never been enough if you are wrong about the price. Rob Go at NextView rounds things out with a sharp framework breaking down the four approaches to early stage venture and what the real failure mode looks like for each. Three great reads with a lot to unpack across all of them.

This week on The Learning Corner, Mia and Charles unpack the resignation letter of an Anthropic safety researcher who walked away from one of the most valuable companies in the world, alongside Anthropic quietly walking back its flagship safety commitment. They then dig into Jack Dorsey's decision to cut Block's workforce nearly in half, not because the business is struggling, but because he believes AI has fundamentally changed what it means to build a company. They close with a sharp LinkedIn post from Euclid Ventures GP Nic Poulos making the case that the venture industry has developed a consensus addiction, and the return data to back it up is hard to ignore.

This week we dig into the viral Citrini Research "2028 Global Intelligence Crisis" memo, a scenario written from the future that rattled public markets and sparked a real conversation about what happens when AI works exactly as promised. We also highlight a great StrictlyVC Download episode with Stacy Brown-Philpot, founder of Cherryrock Capital, on what it takes to win at Series A in today's market. We close with a LinkedIn post from Alex Pall of Mantis Venture Capital on 2021-vintage founders hitting a first-principles reset in the age of AI.

This week we're joined by a special guest, Ariana Ferwerda, CEO and co-founder of Halfdays, a women's ski and outdoor wear brand, to discuss her piece on why her team will always be in the office and the inflection points that brought them from remote to fully in-person. Then, Charles and Mia dig into Kanyi Maqubela's "Bubbles and Balloons," which argues the 2021 market wasn't a bubble but a slow deflation, and whether the AI boom is simply re-inflating the same balloon. They close with Will Manidis' "Tool Shaped Objects," a sharp response to viral AI doom content that challenges whether the current AI infrastructure boom is producing real productivity or just the feeling of it.

This week on The Learning Corner, Mia and Charles unpack why AI is making engineers more productive but more exhausted, exploring the hidden cost of constant tool evaluation and "thinking atrophy." They then dig into a provocative argument that venture capital's traditional model is fundamentally breaking down and why the future looks more like permanent capital than ten year fund structures. Finally, they walk through Julia Maltby's tactical guide on building a GP reference list that actually closes LPs, including why your best reference might be a founder from a deal that didn't work out.

This week features Beezer Clarkson from Sapphire Partners discussing her analysis on venture's first substantial firm contraction in 20+ years, with fewer than half of existing VC firms actively investing. We examine Om Malik's critique of access journalism dominating tech media, where velocity trumps substance. Finally, Michael Dempsey's essay argues that VC-backed startups have undergone the same status collapse as investment banking, becoming the unremarkable default path for ambitious talent rather than a signal of contrarian vision.

This week we discuss Jason Lemkin's analysis of Brex's $5.15B exit and how hubristic fundraising creates impossible success benchmarks, Auren Hoffman's controversial claim that VCs overpromise their ability to help with hiring in today's AI talent war, and Saanya Ojha's defense of SaaS business models against claims of obsolescence. We explore why massive valuations attract mercenaries over missionaries, whether venture capital firms can truly move the needle on talent acquisition, and how vendor fatigue and data moats protect incumbents from AI disruption. Join us as we challenge conventional wisdom on fundraising strategy, investor value-add, and the future of enterprise software.

This week we discuss Sequoia's groundbreaking decision to invest in Anthropic despite existing stakes in OpenAI and xAI, challenging traditional VC taboos around backing competitors. We explore why pre-seed fundraising has fundamentally changed, with most founders now needing $300K ARR just to get meetings. Finally, we examine a candid reflection from an ex-Amazon employee who regretted leaving FAANG for a startup, highlighting the mission-driven intensity required to succeed in early-stage companies.

This week we explore David Haber's framework on why most investors run funds but few build lasting firms, and what it takes to create a defensible moat in venture capital. Sarah Guo reminds us that relevance decays without constant market engagement, especially in the compressed timelines of AI companies where first principles thinking matters more than inherited pattern matching. We close with Michael Dempsey's piece on the privilege of belief in early-stage investing, examining what steadfast conviction looks like when operating with incomplete or no data. From institutional durability to personal conviction, we unpack what separates temporary success from lasting impact in venture.

This week we unpack the viral AI food delivery hoax that fooled thousands on Reddit and what it reveals about misinformation in 2026. We discuss the crisis of career fulfillment in tech as layoffs continue while startups struggle to find talent. Finally, Roger Ehrenberg shares an honest take on the insecurities every venture investor faces and why playing your own game matters more than ever.

This end‑of‑year episode of The Learning Corner by Precursor brings together our first podcast guests, Hunter Walk (Homebrew, Screendoor) and Peter Walker (Carta) for a wide‑ranging conversation on how venture capital has evolved in 2025. We unpack ballooning early‑stage valuations, the distortion created by media attention, and how few companies actually sit at the top of the market. We also discuss today’s exit landscape, growing pressure for liquidity, the rise of secondaries, and what recent IPOs and M&A activity mean for funds of different sizes. We close by examining what early‑stage venture really looks like today, from changing GP‑LP dynamics to faster paths to scale and the push to build more with smaller teams.

In this week’s episode, we break down Ben Casnocha’s shift in thinking on whether good ideas matter more than good founders, Aaron Harris’ challenge to the “missionary founder” trope, and Jonathon Ready’s exploration of why AI is getting eye rolls from engineers in Seattle. We discuss how these takes intersect with early-stage investing and what signals we should be looking for as the startup landscape keeps evolving.

In this episode of The Learning Corner by Precursor, Mia Farnham is joined by Charles Hudson and Missy Martin to unpack three timely reads. First, we explore Charles Schwab’s $660M acquisition of Forge Global and what it signals about the mainstreaming of alternative assets for retail investors. Next, we dig into TechCrunch’s coverage of “venture zombies” and the rise of “hold forever” investors acquiring stalled startups. Finally, we reflect on a powerful essay by Sinéad O’Sullivan that reframes burnout as an identity collapse—not a productivity issue—and what that means for founders, operators, and investors alike.

This week on The Learning Corner by Precursor, Mia Farnham and Charles Hudson dive into three thought-provoking reads shaking up the venture ecosystem: First, we break down Ben Zises’ take on why consistent founder updates may signal future success — and why that doesn’t always align with our own portfolio data. Next, we explore M.G. Siegler’s “Hey, There’s a Bubble” and what today’s AI funding frenzy reveals about hype cycles, economic dissonance, and the uneasy optimism baked into bubbles. Finally, we unpack Kyle Harrison’s “Build What’s Fundable” and its take on YC’s shifting tone, consensus capital, and the rising prevalence of ‘slop startups’—and what it means for early-stage builders trying to think independently in a world that rewards conformity.

We kick things off with a short but sharp post from Ben Choi reminding us that “VC is hard”—not just because of performance pressure, but because navigating fund dynamics and relationships is a game in itself. Then we dive into a provocative piece by Stefano Bernardi questioning the obsession with concentrated funds, and unpack why diversification often outperforms even the best instincts. Finally, we revisit the Cluely hype cycle, with CEO Roy Lee reflecting on the downside of viral traction and what happens when the market stops being impressed.

This week on The Learning Corner, Mia and Charles break down four of the most thought-provoking reads shaping the tech and venture ecosystem right now: • Unsafe SAFEs in the Age of AI: Jason Lemkin calls out a troubling new trend where founders in hyped AI deals walk away with investor cash—without building a thing. • 15 Charts That Explain How Tech and Venture Are Changing in 2025: Ruben Dominguez drops a chart-heavy update covering everything from AI app churn to why so many junior VCs leave the industry. • Oreo-maker Mondelez is using GenAI to slash marketing costs: Mondelez teams up with Publicis and Accenture to roll out a $40M AI tool, cutting costs—and possibly creativity—in CPG advertising. • Sequoia Names New Co-Leads as Roelof Botha Steps Down: With Roelof Botha stepping down, Sequoia is signaling a new chapter

First up, we dive into “Beyond True or False: Teaching Students to Interrogate AI Unreliability”, a Substack by Nick Potkalitsky, which proposes a new framework—borrowed from literary theory—for teaching students to critically evaluate AI-generated content. We discuss how this lens can help people move beyond simple trust/distrust binaries and become better co-creators with AI. Then, we explore “Is AI the New Shadow Bank? (Yes…)”, a piece that draws parallels between today’s AI economy and the pre-2008 shadow banking system. Instead of mortgage-backed securities, today’s collateral is GPU access, compute contracts, and foundation models—and we ask: is the real innovation the credit system AI is built on?

We’re digging into three thought-provoking topics shaping the current startup and investing landscape: 🧼 Coterie’s $650M Exit & the Return of DTC M&A – We break down Brian Sugar’s “One Brand is Luck, Two is Strategy,” and why pure-play DTC brands with strong economics and customer devotion are back on the radar for modern acquirers. ⚡️ Iteration Speed & the Path to Series A – Hadley Harris of Eniac Ventures shares why iteration speed is the best predictor of Series A readiness, and how founders can prioritize feedback loops and kill ideas quickly to increase their odds. 💸 Secondaries in Term Sheets – A recent Axios note suggests Menlo Ventures is now pre-negotiating secondary sale conditions into early-stage term sheets. We unpack what this means, why it matters, and whether this becomes a new norm.

This week on The Learning Corner, Mia and Charles discuss Goldman Sachs’ acquisition of Industry Ventures and what it signals for the venture ecosystem, Maria Gonzalez Blanch’s take on “brain obesity” in the age of AI, and a recent post from Erica Wenger about friends investing in friends. We break down what it means for community, trust, and evolving norms in tech.

In this week’s episode of The Learning Corner by Precursor, Mia and Charles explore three themes shaping today’s startup and VC landscape: The Three Lanes of Modern Venture – What type of fund are you really building? John Vrionis lays out three distinct approaches VCs are taking today. When Great People Leave – How do strong leaders navigate inevitable departures? Lessons from Taps Notes on leadership, transitions, and grace. The Solo Founder Debate – If solo founders can succeed, why don’t more VCs back them? Itamar Novick shares thoughts on survivorship bias and what makes solo builders special. Tune in for a thoughtful discussion on the shifting expectations of investors, how to support teams through change, and why the solo founder conversation is evolving fast.

We start with a Fortune article on how founders are using “creative accounting” to boost ARR, once the gold standard for SaaS success and now a much murkier metric in the AI era. What used to be a reliable sign of recurring revenue has drifted into “vibe revenue,” and we talk about what that means for investors and founders trying to benchmark growth. Next, we dive into Jasmine Sun’s blog on Silicon Valley’s cultural lexicon, “high agency,” “NPCs,” “996,” “taste,” and “decel/doomer.” These memes reflect deeper anxieties about meaning, ambition, and the pressure to win in today’s AI-driven world. We share our reactions to which of these terms resonate most, and what they say about tech culture right now. Finally, we turn to the Wall Street Journal’s coverage of how debt is fueling the AI boom. With Oracle, CoreWeave, and others leveraging massive loans to secure infrastructure and power, we discuss whether this strategy is a smart bet on future demand or a dangerous setup for a bubble if growth slows.

We explore Jerry Neumann's "AI Will Not Make You Rich," which argues that transformative technologies like AI may not deliver lasting competitive advantages, using economist Carlota Perez's tech wave framework to examine whether we're in the "frenzy" or "irruption" stage of AI development. Next, we dive into the curious disconnect in consumer tech markets, where VC funding has dropped from 15% to 5% of venture dollars despite consumer companies significantly outperforming the Nasdaq-100. Charles discusses why Precursor remains bullish on out-of-favor sectors and whether AI-powered personalization could revive investor interest in consumer tech. Finally, we examine Nvidia's $100 billion investment in OpenAI and what this reveals about the interconnected web between chip makers, data centers, and AI companies.

This week on The Learning Corner, Mia and Charles explore three compelling venture capital topics: (1) Micah Rosenbloom's comparison of VCs to "cockroaches" and why fund consolidation remains unlikely despite market pressures, (2) the emerging trend of software engineers being paid to fix AI-generated "vibe-coded" projects that need human expertise to become functional, and (3) new Pitchbook data revealing how serial founders maintain a significant 2-3x fundraising advantage over first-time entrepreneurs in today's cautious investment climate.

First, we dive into “We Have Met the Enemy and He Is Us” by Euclid Ventures, which explores how venture capital is drifting from its roots as a market for independent thinkers. Next up is “The AI Productivity Paradox” from Sequoia’s Inference, which explores why widespread AI adoption hasn’t translated into real productivity gains. Lastly, we break down the recent Reuters story about Anthropic’s $1.5 billion settlement in a landmark author copyright case. Soon after we recorded, the Judge stepped in and rejected the settlement.

This week on The Learning Corner by Precursor, Mia and Charles dive into three major topics shaping the future of AI, startups, and software economics: Taco Bell’s Voice AI Troubles: A WSJ piece reveals just how glitchy the chain’s drive-thru AI rollout has been—and why they’re rethinking it entirely. Charles shares why these failures might actually be a good sign of progress and what’s at stake when AI misfires in higher-risk industries like healthcare and finance. The Myth of “Non-Consensus” Investing: A provocative thread from Martin Casado questions whether early-stage investors are placing too much faith in being contrarian. Is alpha really found in non-consensus bets—or does follow-on capital always chase what’s hot? The Software Margin Cliff: A brilliant Substack from Sam Schillace suggests that AI is driving software economics off a cliff. As inference costs grow and traditional SaaS margins shrink, the industry may look more like manufacturing than the creative playground it used to be.

We kick off the episode by discussing Rebecca Kaden’s essay, “Hire the Experimenters”, which argues that speed, adaptability, and creative risk-taking are more valuable than traditional credentials in today’s AI-driven startup world. We explore how this mindset shift is reshaping what “great hiring” looks like for early teams — and whether that flexibility can translate into long-term defensibility. Then we break down the MIT report on generative AI deployment across enterprise — and the not-so-great news that 95% of pilots are failing. Charles shares his opinion that these "failures" are signs of future success. Finally, we look at Peter Walker’s data-packed breakdown of bubble-era valuations reappearing at Seed and Series A. With median valuations at the 99th percentile soaring past $160M for Seed and $700M+ for Series A, we ask: what does this mean for startups trying to raise now, and how do you tell signal from noise?

This week on The Learning Corner, we dig into the changing dynamics of talent, regulation, and hiring in venture. We start with the San Francisco Standard’s piece on why founding engineers are suddenly the most coveted hire in Silicon Valley, and how the expectations for technical co-founders have shifted post-AI. Next, we explore Platformer News’s report on the political crackdown on AI therapy bots. As states move to regulate AI mental health tools, we debate where the line should be drawn between helpful support and unlicensed therapy, and what it means for builders in this space. Finally, we unpack Amplify Partners’ post on why most VC firms hire associates wrong. Charles shares what he’s learned from hiring across multiple funds at Precursor, and we discuss what it looks like to truly invest in future partners from the earliest stages of their careers.

This week on The Learning Corner, we explore three perspectives shaping how startups get built and funded in today’s environment. We start with Aditya Agarwal’s LinkedIn post on the two competing startup modalities in Silicon Valley: those building in legible, fast-moving spaces where speed is the moat, versus those tackling illegible, contrarian ideas that take longer to crystallize. We discuss how investors say they want one, but often fund the other. Next, we dive into Rosie Bradbury’s PitchBook piece on whether VCs are truly independent thinkers. We break down new data showing how frequently top firms co-invest, and what it means for the myth of proprietary deal flow. Finally, we unpack the New York Times’ coverage of the wave of 20-something founders flooding San Francisco to build AI startups. From teen dropouts to meme marketing, we reflect on how this boom compares to past cycles and what might be different this time around.

This week on The Learning Corner, we dive into three sharp pieces exploring the evolving shape of venture. We kick off with Elad Gil’s take on why AI markets are finally gaining clarity, with key winners emerging across foundational models, code, legal, and customer experience. We ask: can new entrants still break through if they’re not one of the early, well-capitalized names? Next, we unpack Auren Hoffman’s argument that data businesses, while solid, often aren’t built for venture scale. As more investors rethink their appetite for pure-play DaaS models, we explore where data companies can still attract VC interest, especially as AI blurs the line between tooling and infrastructure. Lastly, we discuss Tomasz Tunguz’s piece on why seed rounds are getting larger even as startups shrink. From inflated pre-seed valuations to AI-powered team efficiency, we reflect on the new expectations being placed on early-stage companies and where those pressures might lead next.