Marc Andreessen on Inventing The Internet Browser (How We all Access the Internet)

Mark Andreessen's Background:

• Grew up in rural Wisconsin (population ~1,309) in a lower-middle-class farming family without indoor plumbing or gas heat for much of his childhood.
• Lacked exposure to the internet until university; first encountered computing via cassette tapes (requiring manual fast-forwarding) and low-bandwidth BBS systems.
• Attended the University of Illinois in 1989, arriving at a campus wired for the internet with access to $25 million Cray supercomputers.

The State of the Internet (Pre-1994):

• Cultural Consensus: The overwhelming expectation was that the internet would never be a business; commercial activity was banned under the NSF Net's "Acceptable Use Policy."
• Competitor Alternatives: Major corporations (Microsoft, Apple, AOL) were building proprietary "Information Superhighways" and interactive television systems to capture user traffic, believing the open internet would fail.
• Paul Baran's Legacy: The internet was founded on Paul Baran's heretical 1964 concept of packet-switching (designed to survive nuclear strikes), which AT&T rejected as "dumb" and "impractical" for 20 years.
• Demographics: Users were limited to ~1 million scientists and engineers; the network was text-based, slow, and dominated by technical experts.

Development of Mosaic:

• The Team: Andreessen, Eric Bina, and colleagues at NCSA built Mosaic in 1992–1993, initially working off-book at night to avoid academic bureaucracy.
• Design Philosophy:
  • Heretical Assumption: They designed for a future where every user had a graphical interface (Mac/Unix/Windows) and broadband connectivity, refusing to compromise for older hardware.
  • Text-Based Protocols: Intentionally chose inefficient, human-readable text protocols (HTML, HTTP) over binary ones to maximize openness and allow "View Source" to empower new developers.
  • Inclusivity: Decided to add images (the "image tag") despite strong opposition from the community who feared mass-market content would degrade the scientific nature of the web.
• Launch: Released version 0.9 in late 1992 and 1.0 in spring 1993; became the first unified graphical interface for Gopher, FTP, and News.

Commercialization and Conflict:

• Market Ignorance: Andreessen initially had no knowledge of venture capital, assuming the product was only for academic use.
• Media Skepticism: Early reports from the Wall Street Journal and Wired magazine dismissed the internet as unimportant; the first major New York Times story incorrectly credited Andreessen's superiors rather than the students who built it.
• University Dispute: The University of Illinois licensed Mosaic's source code to a startup called Spyglass.
• Netscape Formation: Jim Clark recruited Andreessen to found Netscape in 1994 to build a commercial browser, deliberately not using the original Mosaic source code.
• Legal Settlement: Spyglass and the University harassed Netscape's sales by threatening to sue. Netscape preemptively sued for tortious interference, settling for a cash payment that cost the University an estimated $3 billion in lost future valuation and philanthropy.

The Microsoft & Spyglass Collapse:

• The Deal: Microsoft licensed Mosaic's code from Spyglass for a ~$1 million flat fee to power Internet Explorer.
• The Blow: Microsoft announced Internet Explorer as a free add-on to Windows, instantly destroying Spyglass's business model and leaving them with no revenue stream.

Forward-Looking Themes:

• Open vs. Closed: The transcript draws a direct parallel between the early internet struggle and the current AI debate, warning that big tech is pushing for "regulatory capture" and closed systems, repeating the mistakes of the 1990s.
• Expert Incompetence: Andreessen cites Phil Tetlock's research showing that credentialed experts were "less than random" at predicting the internet's success, highlighting the systemic lack of accountability for expert predictions.
• Hyperstition: The Mosaic team operated on the belief that building a superior user interface for the future would cause the necessary infrastructure (broadband, GUIs) to emerge.