This interview covers Benchmark's unique investment philosophy and AI opportunities. They believe smaller funds force real partnerships, not just capital. They see AI as a distribution shift (selling 'work' not software) and gaming as a golden age thanks to AI tools. Key holdings: Wildlife Studios (Victor's game company, 4B downloads, no VC), Brex (Victor on board, praised as 'closest to co-founder'), Telegram (~30 employees, near 1B users).
Benchmark partners Peter Fenton and Victor Lazarte discuss the core driving forces behind great entrepreneurs and Benchmark's unique operational philosophy. Victor, formerly a co-founder of Wildlife Studios, which he built into one of the world's largest independent mobile game companies, later tran
Guest identity: Benchmark partners Peter Fenton (long-term partner) and Victor Lazarte (new partner, former co-founder and CEO of Wildlife Studios, which he built into one of the world's largest independent mobile gaming companies). Main theme of this episode: Benchmark's unique operating philosophy, the core driving forces behind great entrepreneurs, and new investment opportunities in the AI era. The most weighty judgment in the entire video: Peter Fenton believes that Benchmark's "smaller fund" model (intentionally reducing the fund size to half of the suggested amount) maintains discipline and focus by limiting scale, thereby preventing capital relationships from replacing genuine partnerships.
Peter Fenton argues that Benchmark deliberately limits its fund size to half the recommended amount, forcing itself to build relationships in non-monetary forms and preventing capital from replacing true partnership. He points out that venture capital was originally a "hack" from the oil exploration model—investors use capital to buy relationships, but "once we invest, a bunch of friends you'll never meet enter your cap table; they run pension funds, will never help you, yet own more shares than your VP of Sales." By capping fund size, Benchmark "does not buy relationships but earns them." Data shows that, on average, Benchmark's portfolio companies see six to seven times value appreciation for their founders. Peter emphasizes, "Larger funds dilute purpose and increase pressure on financial transactions."
Victor Lazarte adds that Benchmark makes only about 10 investments per year, typically in Series A, and partners own an average of zero vacation homes. Each partner ultimately holds roughly 15 basis points of carry (using the Wildlife investment as an example: 1% of a 3% equity stake belongs to Benchmark, then evenly split among six partners). Victor believes this model forces partners to realize that "if you optimize for the short term, you write many checks, but after writing those checks, you have an incentive to spend as little time as possible with founders." Benchmark does the opposite: "Once we make a commitment, the average relationship lasts ten years."
Peter Fenton introduced a "three-motivation model" inspired by Paul Conte that directly maps to entrepreneurs: competitive drive (winning, outdoing others), pleasure drive (pursuing sensory gratification, which becomes habitual), and generative drive (creating, serving, contributing). He argues that the generative drive is the cleanest fuel, while the pleasure drive is the most dangerous—because it "feels so good." In successful entrepreneurs, Peter observed that when founders begin to personally consume that wealth, "historically, it has severely eroded their ability to be great founders and leaders."
Victor Lazarte shared his self-discipline at Wildlife: spending only 1% of the amount deposited in the bank account in the first year. He believes this cultivates a discipline, and "one of the core advantages of Silicon Valley is not that spending a lot of money is cool, but that having great ideas is cool." Victor emphasized that environmental culture is critical—South America lacks the collective reinforcement mechanism found in Silicon Valley that "pulls you back toward the generative drive." He cited data: Wildlife games have been downloaded 4 billion times, with over 1 billion people playing them, yet the company paid tens of millions of dollars in dividends in its early years, and this wealth temptation is a challenge "capable of consuming" many people.
Peter added that this "life's work" mindset needs to be "de-oxidized"—when founders start from zero, after accumulating external rewards, fame, and wealth, they often become "covered by the company's bureaucracy, the traps of wealth, and the distractions of success." He referenced James Carse's concept of Finite and Infinite Games, arguing that a company in an infinite game "is an expression of what you want to contribute to the world, you let these talents work through you, rather than grasping them and using them solely to defeat others," which creates an essential distinction between "the joy of pursuit vs. the pursuit of joy."
Victor Lazarte argues that AI is similar to the internet, not cryptocurrency, because "the fact that computers can speak natural language is a big deal." He notes that AI's transformative potential lies in changing how humans interact with computers — shifting from "looking at a screen and doing this action" to "talking as if speaking to another person." He is particularly bullish on the messaging app space: "Today's messaging experience is very unoptimized — when you communicate in real time, it's more engaging and effective than text messages." He predicts the emergence of a founder "under 30 years old" who will use generative AI technology to create a "deeply missed human connection experience."
Peter Fenton further distinguishes between "technology disruption" and "distribution disruption" in AI. He believes that most people see AI as technology disruption (favoring incumbent giants), but Peter argues there are two types of distribution disruption: first, demand disruption (every company asking "how does this apply to my business"), and second, pricing disruption — shifting from "selling seats" to "selling work" — buying OPEX rather than buying tools. He predicts the rise of application companies that "sell work" instead of "selling licenses": "If you're buying work, you're watching where your operating expenses go — a completely different framework."
Peter notes that Benchmark has not invested in any foundation model companies, a choice consistent with its investment philosophy — favoring teams of fewer than 10 people with the potential to reach 1 billion active users. He cites data: "Over 1 million developers are integrating with Transformer models, which will become the breeding ground for the next 3-5 companies with a $500 billion market cap." He warns that this prediction could be "completely wrong" — "This could be a repeat of cryptocurrency? We don't need to know the answer; we just need to partner with the founders driving the greatest generative force."
Victor Lazarte argues that the gaming industry is entering a "golden age," as AI tools transform game creation from a "technical challenge" into an "imagination challenge." He draws a historical analogy: "In the early days of books, the core advantage was having a good printing press – you hired authors and paid them a small fee. When printing became commoditized, more people could write, and ultimately you got a better selection of books." Similarly, AI tools (such as MidJourney) will make "turning your imagination into reality the easy part, while imagining great experiences becomes the hard part."
Victor emphasizes that the most successful games are often "evolutions" of existing good mechanics. He cites examples: League of Legends is an evolution of Dota (free-to-play), Monopoly Go is an evolution of Coin Master (more accessible). The core pattern is that "humans enjoy the same game mechanics for a long time – think about chess and poker, which have existed for thousands of years." He predicts that the business model for games will remain "free-to-play," because "the experience of trying before committing is excellent for users." However, the real innovation lies in the fact that "tools have become so good that creating a persistently engaging game becomes the core challenge."
| Position | Guest Attitude | Key Data |
|---|---|---|
| Wildlife Studios | Bullish (Victor built the company from scratch) | 4 billion downloads, 1 billion+ players, hundreds of millions in annual revenue, no VC capital used |
| Brex | Bullish (Victor serves on the board) | Founders call Victor "the closest person to a co-founder outside the team" |
| Uber | Neutral (risk warning case) | Benchmark once faced major governance challenges |
| Neutral (historical case) | Early investment, Benchmark partner was part of a 10-person team | |
| Snapchat | Neutral (historical case) | Early investment |
| Discord | Neutral (historical case) | Startup reborn from failure |
| Telegram | Bullish (as a role model) | Nearly 1 billion MAU, only 30 employees, founder self-funded |
| Confluent | Bullish (Eric's case) | Collaborated for over 5 years |
| Amplitude | Bullish (Eric's case) | Collaborated for over 5 years |
| Benchling | Bullish (Eric's case) | Collaborated for over 5 years |
| Airtable | Neutral (mentioned) | - |
| Figma | Neutral (mentioned) | - |
| Scale AI | Neutral (mentioned) | - |
| Monopoly Go | Neutral (mentioned as a success case) | Currently the world's largest mobile game |
| Coin Master | Neutral (mentioned as a comparison case) | Early successful game |
1. Benchmark's "smaller fund" model is essentially an adverse selection of "relationship tax." Peter Fenton argues that the larger the fund, the easier it is to use capital to purchase relationships rather than build them, which in turn reduces founder loyalty and the depth of trust between investors and founders. Data supports this: Benchmark's average portfolio company founder experiences six to seven times value creation, and founders only experience "significantly less dilution."
2. Three-drive model: competitive drive, pleasure drive, generative drive. Peter Fenton maps Paul Conte's three-drive framework directly onto entrepreneurship: the generative drive (creation, service, contribution) is the cleanest fuel, the competitive drive is suitable for sprints but unsustainable, and the pleasure drive (habituation) is the most dangerous—"because it feels so good."
3. AI's "distribution disruption" rather than "technology disruption" will create new platform opportunities. Peter Fenton believes that most people wrongly view AI as a technology disruption (benefiting incumbents), but there are two types of distribution disruption: demand disruption (every company asks "how to apply") and pricing disruption (shifting from "selling seats" to "selling work"). Falsification condition: if incumbents successfully internalize AI applications, this thesis fails.
4. The "life's work" needs to be "deoxidized." Peter Fenton argues that successful founders undergo a process of "being covered by bureaucracy, wealth traps, and distractions from success," and the investor's core job is "to deoxidize, polish, and bring the most creative abilities to the surface"—which requires "deep caring" rather than "transactional relationships."
5. The new golden age of the gaming industry comes from "tool commoditization + imagination differentiation." Victor Lazarte explains via an analogy with book printing: when printing technology became a commodity, more people could write, and ultimately better book selection resulted. Similarly, AI tools will make "imagining a great experience the hard part, and executing the technology the easy part."
6. Core vulnerability of the Benchmark model: survival conditions without an "alpha." Peter Fenton uses the Chimp Empire analogy to explain: Benchmark's equal partnership model resembles the "Western troop" (30-40 chimpanzees, egalitarian, sharing food), while the "Central troop" (100+ chimpanzees, alpha male dominated) requires massive energy to maintain status. Benchmark's vulnerability lies in "the lack of hierarchy can lead to complacency"—if the firm does not internalize a "sense of excellence," it will quickly become irrelevant.
7. Buying "work" rather than "licenses"—the new business model for AI applications. Victor Lazarte predicts that AI will give rise to application companies that "sell jobs"—users purchase "reductions in operating expenses" rather than "tool licenses," which will fundamentally change software pricing and distribution logic.
8. Human flourishing and mental health—the biggest opportunity for yet-to-be-founded companies. Peter Fenton believes that "one of the major yet-to-be-founded companies will solve the pathology of people feeling 'bad' after spending 20-30-40 minutes on TikTok or Instagram"—he cites the historically highest suicide rate among Gen Z, calling it the biggest topic "the industry has not yet fully discussed."