The a16z Show

Replay: How the Internet Happened | Chris Dixon & Brian McCullough (2018)

· Summarized

Share this episode
Anyone with this link can open it.
Share Link
Start at 0:00

Key takeaways

  • Technologies dismissed as bad ideas are often ideas whose infrastructure, market, or users are not ready yet; founders should distinguish permanent flaws from timing problems.
  • The open web prevailed over centralized information-superhighway plans because good-enough, permissionless tools reached users and developers before more polished systems were ready.
  • eBay’s lasting contributions were teaching strangers to trust one another, popularizing scalable reputation systems, and proving that a platform need not own the underlying goods or infrastructure.
  • Napster’s central insight was that people wanted convenient, immediate access to media; Spotify later validated that value proposition with a simple service at a reasonable price.
  • Google succeeded not only through better search but by pairing it with relevant auction-based advertising, creating value for users, advertisers, and the search company simultaneously.
  • The next major technology wave is likely to emerge from hobbyists and developers who are willing to explore for years without a clear short-term business case.

Summary

Why McCullough wrote a history from Netscape to the iPhone

Chris Dixon praises Brian McCullough’s How the Internet Happened as accurate, entertaining, and potentially a canonical history for younger technologists who have never known a world without the internet. McCullough says he wanted to explain to ordinary people—not just technologists—who built the systems now embedded in daily life, why they made particular choices, and how they did it. The book covers roughly 1993, when Netscape brought the web into public view, through 2007, when Apple announced the iPhone. Both argue that understanding this history supplies pattern recognition: today’s builders can learn not only what worked, but why apparently good ideas failed because they arrived too early.

The “bad ideas” that were simply too early

Dixon recalls a 2005 eBay card game, apparently called Dot Bomb, that treated internet grocery delivery, internet money, and similar ventures as jokes. He says many of today’s hottest technology companies grew from precisely those supposedly foolish ideas. McCullough connects this to Marc Andreessen’s formulation that there are no bad ideas, only ideas that are too early. Grocery delivery, cloud computing, cloud storage, and online pet-food sales failed when infrastructure, bandwidth, logistics, or consumer behavior were insufficient, but later became viable. The dot-com period therefore resembles a technological Cambrian explosion followed by an asteroid impact: founders must learn both how to build and how to survive industry-wide collapses.

How the open web beat the information superhighway

Dixon and McCullough revisit the early contest between a decentralized web and centralized “information superhighway” visions associated with Comcast, Disney, Microsoft, Barry Diller, and others. Those companies expected broadband, a large pipe, and a box beside the television to deliver a controlled, top-down experience. McCullough clarifies that they were not wrong about broadband’s eventual importance; they were wrong to assume they had time to perfect the system. Mosaic, home-built websites, Yahoo, and HTML offered a rough but usable alternative. Berners-Lee’s simple HTML, DNS, and the web’s permissionless architecture let hobbyists and universities create on nearly equal terms with Disney. Good-enough technology won because it reached users and developers sooner.

eBay taught strangers to trust one another online

McCullough argues that eBay may be the most influential and genuinely internet-native company of the early era, even though it is not usually grouped with Google, Amazon, or Facebook. First, eBay trained ordinary people to trust faceless strangers across the country, at a time when people still wondered whether anyone would put a credit card online. Second, Pierre Omidyar’s user-rating system showed that self-organized reputation could regulate a mass marketplace without central authority; Omidyar created it partly because he did not want to settle disputes himself. Finally, eBay proved that a company could create value simply by enabling users to interact, without owning factories, trucks, inventory, or the goods being exchanged.

Napster understood that convenience—not only piracy—was the product

McCullough says Napster’s founders were publicly describing, as early as 1998 and 1999, a future of unlimited selection and instant access. Their failure was not that the underlying value proposition was wrong, but that they challenged an industry with powerful legal defenses and could not attract the necessary talent or monetize the service. Dixon and McCullough frame Napster as a three-act story: it was condemned as piracy, shut down as though the problem had been solved, and later vindicated when Spotify demonstrated that people would pay for simple, reasonably priced access. Had record companies tested Napster instead of fighting it, the company might have pioneered music and then video streaming before Spotify or Netflix.

Google combined superior search with an accidental advertising fit

The hosts describe Google’s two major breakthroughs as solving search and finding a business model that matched it. GoTo, later Overture, invented auction-based search advertising, but Google made the system more effective through relevance and quality scoring. An advertiser could pay less per click when an ad was more relevant; Google could earn more from a lower bid that generated many clicks than from a high bid that generated few. Users also benefited because relevant paid results supplemented strong organic results. This contradicted the portal model, in which companies surrounded mediocre search with banner ads and celebrity content because a truly useful search engine would send visitors away. Google initially sold expensive enterprise “Google boxes,” since investors doubted consumer search could make money.

The dot-com bubble mixed experimentation with a greater-fool race

McCullough says early internet entrepreneurs were exploring several unknowns at once: the right product, business model, distribution system, and proof that commerce itself could work online. The first wave included serious companies and well-intentioned failures; later, around the height of the boom, cynical strategies emerged in which venture capitalists aimed merely to take companies public and sell before others did. Wall Street money and baby boomers reaching peak earning years added fuel to the longest bull market, which had begun in 1982. McCullough does not treat every failed founder as foolish: no one could yet see the whole “elephant,” and much of the outcome depended on timing and chance.

From Web 2.0 to mobile, each platform unlocked a different kind of participation

McCullough distinguishes the original web’s commerce-heavy expansion from Web 2.0’s bidirectional model, in which Wikipedia, Facebook, and other services let users create rather than merely consume. Dixon observes that mobile arrived alongside social at an unusually consequential moment: Facebook opened registration broadly while the iPhone appeared within roughly the same 18-month period. The phone was not merely a better screen for consuming social media; its camera, connectivity, and always-available interface made it the ideal device for creating the photos, messages, and other material that social platforms needed. They describe this as an infrastructure-and-application cycle: new infrastructure creates room for a decade of new products, after which the easy opportunities become saturated.

Why the next wave may come from hobbyists rather than incumbents

McCullough thinks the internet was in a lull around 2018: mobile’s low-hanging fruit had been harvested, and founders were still following the “reach a billion users” playbook. Dixon suggests that a new infrastructure layer—cars, virtual reality, blockchain, machine learning, or something unknown—could restart the cycle, while McCullough emphasizes a generational and cultural shift. Large platforms can copy features quickly, as Facebook did with Snapchat-like stories, limiting independent companies before they mature. Both recommend watching hobbyists and developers who build passionately without expecting payment. The Homebrew Computer Club and early web builders invested on ten-year horizons that quarterly businesses cannot, and those apparently impractical experiments often become the next industry.

Show notes

Why do technologies that look like bad ideas in one era become enormous companies in the next? Originally recorded in 2018, Brian McCullough and Chris Dixon trace the companies, failures, and accidents that built the modern internet.

They revisit the battle between the open web and more centralized visions of the “information superhighway,” why eBay helped establish trust between strangers online, what Napster understood about digital media, and how Google found a business model for search.

They also look at the dot-com crash and the ideas that arrived before the technology was ready, from cloud storage to grocery delivery, asking what those cycles can teach founders about spotting the next wave.


Resources:

Follow Brian McCullough on X: https://x.com/brianmcc

Follow Chris Dixon on X: https://x.com/cdixon

Stay Updated:

Find a16z on YouTube: YouTube

Find a16z on X

Find a16z on LinkedIn

Listen to the a16z Show on Spotify

Listen to the a16z Show on Apple Podcasts

Follow our host: https://twitter.com/eriktorenberg

Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures.


Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Follow this show on AudioBrevity and new episodes get summarized for you. Create a free account

Couldn't load the audio. A blocker may be stopping the show's tracking links. Open the file