Inside the Mind of A Famous Investor | Marc Andreessen

KEYPOINTS

• 01

  When you don't understand the present, read backwards. Most 'new' patterns are old.

  Many of Andreessen's assumptions about how the US, politics and society worked were 'upended' between about 2008 and 2015, so he started reading history to see how much of today is actually new. The further back he goes, the more universal the themes look, which he finds 'calming'. Take the claim that a new technology will ruin the next generation of kids: ask whether the same claim was made about earlier technologies, whether it was the same claim, and how it played out. He points to a rich literature in the history and sociology of technology.

• 02

  Society meets new technology in three stages: ignore it, argue against it, then call names.

  He draws this from MIT historian Elting Morison's short book Men, Machines, and Modern Times, written more than 50 years ago, before the internet. First, experts dismiss the new thing. Second, the status quo assembles rational arguments for why it can't work. Third, it 'goes bananas' and accuses the technology's proponents of moral crimes. The reason is that technological change is really a reordering of power and status, and status fights are the most vicious ones humans have. Technologists who think they've built a better mousetrap are surprised to find everyone hates them.

• 03

  The deep conflict today is hierarchies versus networks.

  Between about 1880 and 1939 the West got very good at giant top-down systems: mass manufacturing, mass media, mass government and two-party mass politics, with power exercised in conference rooms, as described in The Organization Man. The internet arrived at near-peak centralization with the opposite architecture: no central node, permissionless, lateral. In 2006, getting an app on a phone meant pitching carriers in a conference room for a place on 'the deck', with opening bids of 40–50% of your revenue. Soviet dissidents risked being shot for owning a mimeograph machine. Now people use the internet both to entrench institutions and to disrupt them, on the left and on the right.

• 04

  The optimistic case is a rebel alliance of internet-educated kids. Every winner becomes an incumbent in about five years.

  Founders now arrive from everywhere, far more knowledgeable than he was at their age because they learned online. The fight is asymmetric: an empire must defend against every attacker, while challengers only need to win once. But a successful startup has about five years before it starts acting like an incumbent, which is rational because it now has something worth defending. Often it's a 26-year-old angry at a 21-year-old. He relays Peter Thiel's harsher view that Google has become anti-tech, but frames it as 'the incumbent temptation' rather than a judgment of any one company.

• 05

  There are two pessimistic cases: 1984, or networks that destroy but can't build.

  In the first, technology becomes a tool of top-down control. Peter Huber's Orwell's Revenge rewrites 1984 with two-way telescreens, and the rebels win. The second is Martin Gurri's thesis in The Revolt of the Public: peer-to-peer networks are great at tearing hierarchies down but haven't proven they can govern, run a democracy or replace banks. That leads to nihilism. Niall Ferguson's The Square and the Tower expresses a similar fear, and Andreessen agrees with the first two-thirds of it and is 'very mad' at the last third.

• 06

  Scar tissue shrinks what you'll explore. In probabilistic fields, 'there are no bad ideas, just timing'.

  Life is touching one hot stove after another: a startup fails, so 'don't work at startups'; you're laid off, so 'never trust big companies'. Your aperture narrows with age. In top-tier venture, about half the companies in any period fail, many simply because they're too early. Webvan failed and food delivery is now huge. He believes that by now almost every failed dot-com idea of the late 1990s has worked. So a16z asks only whether this is the right person and the right time. He cites Annie Duke's warning against 'resulting': in poker, play the next hand the same way, and Kobe Bryant is near the top of the list for missed shots.

• 07

  Judge success by the portfolio, not the individual bet. 'We're in the fluke business.'

  Every top VC in history missed almost all of the great deals of their generation. So a16z judges baskets: a gaming push might back ten studios, and it's fine if one produces 99% of the profit. When a partner's ten investments include one huge hit and nine flops, someone will call the hit a fluke. Andreessen's reply: 'the whole point is to get the flukes.' Not needing 100% means you can take chances. That's true of creative work generally, though not of running a nuclear plant.

• 08

  One partner pulls the trigger: budgets, not votes.

  Some firms require unanimity or a 'council of elders' giving a thumbs up or down. a16z gives partners budgets, and within them they can invest in what they want. In exchange, they must do deep work, keep no secret deals, and present to colleagues who may 'crap on it' without having a veto. A failed investment is acceptable. Missing the category winner because you didn't know it existed is not. The model favors offbeat ideas that fit no box, and it adds speed. It also avoids the 'idea event horizon', the point where veterans say 'TikTok, really?' and stop processing anything new.