Startup stagnation
David Peterson

Walk into a coffee shop in NYC, Berlin, or Tokyo, and you will likely see the same minimalist furniture, the same exposed brick, and the same sans-serif menu. This sameness isn't limited to architecture. In movies, the share of blockbusters that are prequels, sequels, or spin-offs has tripled since 2000. In music, the charts are dominated by catalogs of dead or aging artists owned by private equity firms.

Culture is stuck. It has become an oligopoly where fewer players control more of the market, recycling the same intellectual property.

We are seeing the exact same phenomenon in the startup ecosystem.

Venture capital relies on power laws. It needs massive, non-linear outcomes from high-variance bets to work. But the current crop of startups feels decidedly low-variance. We are drowning in consensus incrementalism. AI for legal and AI for customer support. AI notetakers. AI voice agents. Each category pure consensus. Each category funded to the gills.

Venture capital is designed to hunt for black swans. Why are we filling our portfolios with white ones?

I found a potential answer while listening to Derek Thompson’s podcast Plain English over the weekend. If we overlay the theories explaining cultural stagnation onto the tech sector, a unifying theory emerges. We aren't suffering from a lack of creativity. We are suffering from an excess of optimization.

Reputation optimization

One of the most compelling explanations for why culture has stagnated comes from Adam Mastroianni: we value our lives too much.

Data shows a massive decline in deviance across the board. Teenagers drink less, smoke less, and break fewer rules. Adults commit fewer crimes. Even the number of serial killers has collapsed since the 1980s. Mastroianni argues that as life becomes safer and wealthier, the "value of a statistical life" has risen. And when you have more to lose, you take fewer risks.

This applies to careers as much as it applies to wearing seatbelts.

Twenty years ago, founding a company was a deviant act. It was for misfits who couldn't function in a traditional corporate hierarchy. Today, entrepreneurship is a high-status career path. It is the logical next step after Stanford or ETH and a stint at McKinsey.

The modern founder has a reputation to preserve. Building something truly weird, something illegible that might fail embarrassingly, threatens that asset. A "respectable failure," like a legal AI startup (an obviously good idea!) that gets acqui-hired, preserves optionality. You can still get a job in Big Tech or one of the AI labs afterward.

The rebel has professionalized. But you cannot professionalize deviance. When you remove the risk of social and reputational death, you remove the mechanism that produces outliers.

The curse of feedback

In the 1990s, movie studios didn't have perfect data on what audiences wanted globally. They took bets. Today, technology allows for much tighter feedback loops, and that has changed what we create. We know exactly what worked yesterday, so we make more of it today.

This creates a "competence trap." In science, despite more researchers than ever, experts rate new discoveries as less impressive than older ones. Scientists, overwhelmed by the glut of publications, cluster around safe, established subjects.

Startups are stuck in the same loop. For a decade, ZIRP made money free. This didn't create weirdness; it created noise. To filter that noise, investors retreated to pattern matching. So we codified the act of starting a company. The Lean Startup. YC Startup School. SaaStr and SaaStock. Whether you’re building a marketplace a DevOps tool or enterprise SaaS, we had benchmarks for every stage of growth. We kne exactly what a Series A company should look like.

Now that money is expensive, the pressure to show immediate, legible progress is even higher. Founders are smart; they reverse-engineer the metrics. They build companies designed to get funded, not companies designed to exist. They optimize for probability of success rather than magnitude of success. And investors are just as much to blame. Why do you think founders are fudging revenue growth metrics, annualizing mere weeks of sales data to insane “ARR” numbers? They’re responding to what investors say they want.

The cover band economy

In music, copyright laws and streaming platforms have made the past as accessible as the present. Investors buy rights to old hits because they are safe assets. It makes more financial sense to produce a biopic about Elvis or sample an old hit than to break a new artist.

The tech equivalent is the “X for Y” startup. This isn’t a new phenomenon. We all remember when “Uber for X” startups were a dime a dozen. Now it’s “Cursor for X.” Or the “AI rollup.” Or the “full stack AI-powered services business.” So many of these startups feel like cover bands, playing a slightly different version of a song we all know, hoping to ride the wave and fundraise quickly based on the legibility of a theme we’ve all decided is “smart.”

This environment kills "cults." Mastroianni notes that cults peaked in the 1970s and vanished after 2000. As Peter Thiel wrote in Zero to One, a great startup is effectively a cult: a small group of people bound by a secret or a belief the rest of the world rejects. But secrets are hard to keep in a world of total information transparency. As soon as an idea shows traction, it is cloned. This has become even more true as the startup world has gone mainstream and as the cost of code has gone to zero. Anything that is seen to work is copied endlessly, mined for capital and influence. The marginally weird doesn’t stay weird for long.

A return to weird

A society that minimizes risk is a pleasant place to live. It is safer and cleaner. We are less likely to be murdered or join a dangerous cult. All good things!

But a startup ecosystem that minimizes risk is a disaster.

The market is efficient at pricing known quantities, but it is inefficient at pricing weirdness. That is the reason venture has historically worked. By optimizing for reputation and consensus, we are filling portfolios with ideas that look smart, but won’t change the world.

There are signs the pendulum is swinging back. We’re seeing a resurgence of super technical, deeply weird companies with maverick founders innovating on the edges. It’s a promising sign that a solid portion of that latest YC class are companies that previously would be considered “frontier” tech, but the real outliers are likely operating even further out. At Angular, my partner Gil has started calling this weirdtech: companies that don't fit existing categories and that operate in markets others avoid either because they’re boring or hard or both. These are exactly the areas we unabashedly love.

The advantage of an overcrowded consensus is simple: the field is wide open for things that are actually different. As consensus ideas soak up attention and capital, the founders building in weird markets have zero competition. They have time to be illegible, to iterate without the spotlight and to build credible monopolies. The cover bands will keep playing the hits. But somewhere, quietly, the new music is being written.

The AI-Native Enterprise Playbook
Ten real-time observations on a rapidly evolving playing field

No More Painting by Numbers
It’s the end of the “SaaS playbook.

The Age of Artisanal Software May Finally be Over
Every wave of technological innovation has been catalyzed by the cost of something expensive trending to zero. Now that’s happening to software.

Founders as Experiment Designers
David on why founders should run everything as an experiment.

ENTERPRISE/TECH NEWS

IBM + Confluent. IBM announced it would buy data streaming giant Confluent for $11B. “Confluent offers a platform that helps enterprises manage streams of data in real time, a use case that’s exploded in demand as ever more companies develop and deploy AI products, which require significant back-and-forth processing of data for inferencing. IBM said Confluent will complement its existing data and automation products, as well as improve upon its existing offerings across AI, automation, data, and consulting. The company expects the deal to add to EBITDA and free cash flow in the two years after the deal is closed. This is the latest in a string of deals IBM has struck in recent months as it seeks to capitalize on the AI boom, though at $11 billion, Confluent would be the tech giant’s largest buy in years, following its acquisition of HashiCorp in 2024.”

Going public. Anthropic gears up to go out. "Anthropic, the artificial intelligence startup backed by Alphabet's Google (GOOGL.O), opens new tab and Amazon.com (AMZN.O), opens new tab, has hired the law firm Wilson Sonsini to prepare for an initial public offering that could take place as early as 2026, the Financial Times reported on Tuesday. An IPO would give the company, which operates the AI chatbot Claude, a more efficient way to raise capital and provide leverage for bigger acquisitions through public stock. The move comes as AI adoption gains pace, driven by higher enterprise tech spending and growing investor appetite."

Code red. OpenAI apparently announced a “Code Red” in an internal memo, in response to momentum from Google Gemini and Anthropic (WSJ article, paywall). “The companywide memo is the most decisive indication yet of the pressure OpenAI is facing from competitors that have narrowed the startup’s lead in the AI race. Of particular concern to Altman is Google, which released a new version of its Gemini AI model last month that surpassed OpenAI’s models on industry benchmark tests and sent the search giant’s stock soaring. Gemini’s user base has been climbing since the August release of an image generator, Nano Banana, and Google said monthly active users grew from 450 million in July to 650 million in October. OpenAI is also facing pressure from Anthropic, which is becoming popular among business customers. With OpenAI committed to hundreds of billions of dollars in future data-center investments, concerns about its timeline for turning those investments into meaningful revenues have sent tremors through the stock market in recent weeks. While the company remains private—Chief Financial Officer Sarah Friar said at a Journal event in November that an IPO wasn’t on the immediate horizon—its fortunes are closely bound with those of Nvidia, Microsoft and Oracle, among others."

HOW TO STARTUP

Misalignment. VC Nic Poulos from Euclid Ventures warned against the growing misalignment between VCs and founders. "Raising from a misaligned investor can create lasting, irrevocable issues for a founder. A $400M exit, for example, may be transformational for a founding team—perhaps even the best possible outcome—but a rounding error for a GP who needs $1B+ outcomes to drive returns on big AUMs, is OK with a few zeros, and has the investor rights to veto. On the flip side, small checks can be misaligned too: either because big funds lobbed an option check early and need to buy up to make the position matter, or because a party rounds yielded no single investor with enough skin in the game to truly step up. In short, alignment matters—and it’s worth dissecting the math behind investor incentives to understand how much any given VC will need your company to succeed."

HOW TO VENTURE

GC’s rollup strategy explained. Molly O’Shea released a great in-depth interview with Marc Bhargava, Managing Director at General Catalyst, who went into detail on their roll-ups strategy. "One of the 3 largest venture players with ~$40B AUM, General Catalyst has quietly built a $1.5B AI roll-up engine, investing $100-150M into incubating AI-native companies & acquiring the fragmented services businesses they can transform — a model that sits directly at the intersection of venture creation, operational transformation, & what has traditionally been the realm of private equity."

Sourcing is different now. Charles Hudson argues that VC sourcing has fundamentally changed. "If there was an era when simply having an early signal put you at the front of the line for meeting with a new founder, that era is long over. In addition to early signal, your outreach needs to come with more than just a recognition that you know or suspect the person might be starting a new company - the best founders are discerning and being first or early doesn’t mean what it used to in a world where everyone has their own system for surfacing people who are on the verge of starting new companies."

Lix launches new landing page.

David Peterson quoted on TechCrunch’s article - VCs deploy “kingmaking” strategy to crown AI winners in their infancy.

CruxOCM
Pipeline Simulation Engineer (Remote, Vancouver, Houston)

FalkorDB
Core Developer (Tel Aviv, Hybrid)

Motorica
Account Executive (Remote)