Brendan Eich on Creating JavaScript, the Browser Wars, Firefox and Brave

High-Level Overview

Lex Fridman sits down with Brendan Eich, the creator of JavaScript, co-founder of Mozilla (which built Firefox) and co-founder and CEO of Brave Software, for a three-hour tour through the history of the web. Eich recounts how he wrote the first version of JavaScript in about ten days at Netscape in 1995, why its quirks got frozen in place, how it was standardized as ECMAScript, and how it became fast enough to run Unreal Engine in a browser. He then walks through the browser wars against Microsoft, the birth of Mozilla and Firefox, the rise of Chrome, and why he started Brave to block tracking and rebuild online advertising around user privacy with the Basic Attention Token.

The tone is dense, nerdy and full of insider anecdotes, with Eich's encyclopedic memory for names and dates and Lex's enthusiasm for programming history. The conversation moves chronologically from Eich's physics-student beginnings through JavaScript's evolution and the browser wars, then broadens into privacy economics, censorship, big tech, California and legacy.

The Discussion Explores:

• Eich's path from physics to programming, and his view that science has favored models over experiments
• The origin of JavaScript at Netscape: Scheme ideas, Java-like syntax, and a ten-day sprint
• "Worse is better" and why shipping first beat perfection on the web
• Early mistakes such as the sloppy equality operator, and the lack of garbage collection
• Standardization through ECMA, the long road to ES6, and TypeScript, jQuery and HTML5
• Making JavaScript fast: JITs, asm.js, WebAssembly and why JavaScript became the most popular language
• The browser wars, Microsoft's tactics, the birth of Mozilla and Firefox, and Chrome's arrival
• Brave, tracking and cookies, the Basic Attention Token, user-owned social graphs, and legacy

Deep Dive into Main Ideas

From Physics to Programming

Eich fell into programming helping a high school friend write a BASIC version of the Star Wars Death Star trench run on a Commodore PET. He was a math and science kid who loved HP calculators and studied physics at Santa Clara University, where he dodged Fortran, fell in love with procedural abstraction in Pascal, and admired its creator Niklaus Wirth. He tried to build his own computer and port Unix to it but picked a buggy National Semiconductor chip that never worked, and he switched into math and computer science because the valley was booming while physics looked slow.

Asked about theoretical physics, Eich says the problem is theory over experiment. Mathematics isn't all physical, he argues, and choosing among competing theories on aesthetics is a poor guide. He sees the same problem in economics and epidemiology, where models become policy tools without being tested against reality. Lex relates this to machine learning, which is mostly empirical tricks and data, and Eich says theoretical computer science like P versus NP rarely matters in his practical work, where design is still an art.

How JavaScript Was Born

After seven years at Silicon Graphics and three at the ambitious startup MicroUnity, which Jim Clark told him was doing ten risky things at once, Eich joined Netscape in April 1995, just before its IPO. He had been recruited to put Scheme in the browser, but Netscape had struck a deal with Sun to embed Java. Eich, Marc Andreessen and Bill Joy saw the need for a companion scripting language for designers and amateurs, much as Microsoft paired Visual C++ with Visual Basic. Management told him to make it look like Java, which meant C-style curly braces, Java-like data types and operators. But Eich wanted first-class functions from Scheme, because the browser was single-threaded and needed callbacks, and he borrowed HyperCard's event-handler style, which is why names like onclick and onmouseover exist.

He explains that JavaScript survived while Java died in the browser because Java was stuck in a "plugin prison," confined to a rectangle, while JavaScript ran in the browser's main thread and could interact with the page directly. Early Netscape betas already had developers building single-page apps with local data, prefiguring Ajax. He also recounts the two-way comment-hiding hack that let JavaScript live inside HTML without appearing as gibberish in older browsers, and Microsoft later inventing what became XMLHttpRequest after kicking Java out of Windows.

Ten Days, Worse Is Better, and Regrets

Eich wrote the first version in about ten days, from a Thursday through the following week, by hand-writing a recursive descent parser from compiler knowledge he'd built up in college. He explains Richard Gabriel's essay "worse is better": on networks, moving fast and getting on first wins, like a successful gene sweeping through a population despite flaws. He says he's "way over the fact that early mistakes got frozen by web compatibility, and that the real alternative wasn't a better JavaScript: the alternative was not to do it," which he believes would have meant Visual Basic Script.

His biggest regret came just after the ten days: early internal adopters asked for implicit type conversion in the equality operator so numbers could match string database fields, and he agreed, breaking its equivalence properties. His lesson is that when people ask you to make something sloppy or add a cute feature, the answer should be no, echoing Wirth's line that "the essence of design is leaving things out." He also skipped garbage collection in the rush, so long-running scripts ran out of memory, and the reference counting he added in a hurry created security risks. What he's proudest of is first-class functions, a bit of Scheme that made it in.

Standards, ES6 and the Ecosystem

Netscape took JavaScript to the European standards body ECMA, which had already stood up to Microsoft; since Sun owned the Java trademark, the standard became ECMAScript. Eich describes working with Microsoft's Shon Katzenberger and Sun's Guy Steele, and how the market leader's implementation tends to become the de facto standard. After Netscape was sold to AOL and Microsoft let Internet Explorer stagnate, standardization stalled. An ambitious ES4 was mothballed, but its best ideas, some of which had lived on in Adobe's ActionScript 3, landed in ES6 in 2015, which added features for programming at scale.

Eich likes TypeScript, from Anders Hejlsberg's team at Microsoft, because every JavaScript program is a TypeScript program and developers can add types incrementally as warnings. He credits John Resig's jQuery for showing how verbose the old DOM APIs were, which fed into simpler modern standards like the fetch API and query selectors. He praises his friend Ian Hickson's pragmatic work on HTML5, and notes that Steve Jobs' refusal to allow Flash on the iPhone was the real death blow for Flash, though he admired Flash's timeline animation model.

Making JavaScript Fast

JavaScript performance took off around 2008, when Chrome's V8 and Mozilla's TraceMonkey competed on benchmarks, building on just-in-time compiler ideas from the Smalltalk and Self world; V8's lead, Lars Bak, came from Animorphic, which Sun bought to build HotSpot. At Mozilla, Alon Zakai built a compiler from C and C++ to JavaScript, and Dave Herman and Luke Wagner codified it into asm.js, a typed subset of JavaScript that uses bitwise operators Eich originally copied from Java to mark integers. In 2013 Epic ported Unreal Engine to the browser in four days, running at 30 frames per second with no plugin. That proof led browser makers in 2015 to agree on WebAssembly, a binary format running in the same virtual machine.

Why is JavaScript the most popular language? Eich points to first-class functions, a familiar C-style core, years of listening to developers, and a vibrant standards community. He compares loving JavaScript to the scene in Serenity where Mal tells River she has to love the ship: "you have to love it." Lex compares C++ to Wall Street and JavaScript to WallStreetBets, prompting Eich to discuss oligarchy, Robinhood and what he sees as the hacker ethos of that moment. He even jokes about Gary Bernhardt's talk imagining a future where everything runs on JavaScript.

Advice for Programmers

Eich recommends learning several languages as different tools: APL-family languages for linear algebra, Python for machine learning, a dynamic language for rapid prototyping, and Rust for low-level work. He was Rust's executive sponsor at Mozilla because the team was tired of remote code execution bugs caused by C and C++, and Rust's ownership system provides memory safety and protection from data races. He also admires Go for server-side networking. On whether beginners should start with Python or JavaScript, he says Python, descended from the teaching language ABC, catches more mistakes early, while JavaScript lets beginners commit blunders that go unpunished, such as getting undefined from indexing into a number.

The Browser Wars and the Birth of Firefox

Eich says Internet Explorer won because distribution matters more than anything, and describes Microsoft's hardball tactics: Netscape's head of sales, Ram Shriram, got off the phone ashen after Compaq said Microsoft would pull its Windows license if Compaq shipped Netscape as the default browser. Netscape 4 was late and weak while IE4 was genuinely good. After AOL bought Netscape, Mozilla launched in 1998 as an open-source "escape pod." Eich became acting manager, Mitchell Baker, the lawyer who wrote the Mozilla licenses, became its leader, and when Netscape laid her off she simply showed up to the next community call. He says the weakest Netscape engineers were the ones who disliked Mozilla, while community hires were often better than the median Netscape hire.

A small group including Dave Hyatt and teenage intern Blake Ross stripped Mozilla down to a single browser with tabs and pop-up blocking, which went through the names Phoenix and Firebird before becoming Firefox. Eich overheard a Microsoft employee on a plane say they should have killed Firefox in the cradle by adding tabs and pop-up blocking. Firefox's main revenue came from Google's default search deal, starting in 2004, and Eich told Larry Page in 2005 that Google should build its own browser. Chrome arrived in 2008 with process isolation, famously demoed with a "sad tab" when Flash crashed, and Firefox's share peaked around 2011. He also recounts Mozilla's costly switch of its default to Yahoo in 2014.

Brave, Tracking and the Basic Attention Token

Eich explains how cookies were invented to keep people logged in, but combined with third-party images and scripts they became a way to follow people across sites and build profiles, something nobody intended. He argues browsers should block that by default, and says distinguishing good first-party scripts from tracking requires human and machine learning together, including against fingerprinting and "super cookies." He praises Safari's intelligent tracking prevention and says "privacy is not a unitary good," but a complex, rising wave. He criticizes Google as the new Microsoft in standards bodies and describes its Privacy Sandbox proposals as over-engineered.

Brave blocks ads and trackers, making pages faster and saving battery and data. Its Basic Attention Token, built on Ethereum, aims to reconnect users, publishers and advertisers without middlemen: he cites a Guardian experiment in which about 70% of ad money went to intermediaries. Brave's opt-in private ads are matched by machine learning inside the browser against a daily catalog, so Brave never sees user data, and users receive 70% of the ad revenue, which they can withdraw after identity verification or give to creators. He imagines a Patreon-like creator payment built into the browser for a 3% fee, and a browser that lets you own your social graph across Facebook and Twitter so a ban on one network doesn't cut you off. On censorship he says "the internet routes around censorship," while acknowledging that platforms that refuse all moderation get colonized by terrible people. He jokes that if Brave ever reached 95% share, "I'm going to demand a recount."

California, Mortality and Legacy

Eich is critical of California's taxes, forest management, housing and what he calls corruption, noting friends moving to Texas and Florida, but he credits federal Cold War funding for building Silicon Valley and says its remaining advantage is its network. Asked about mortality, he says he is Roman Catholic and does not fear death. His legacy, he says, has "more to do with my children and and their children," as well as web standards, Firefox and Brave. Firefox proved you can restart a browser market; he says a stubborn minority of lead users, a few percent, can tilt a market even against incumbents. Lex closes with a quote from Jeff Atwood: "any app that can be written in JavaScript will eventually be written in JavaScript."

Conclusion

Brendan Eich's story is a case study in pragmatic engineering: a language rushed out in ten days under commercial pressure, flawed from the start, yet improved for 25 years by listening to developers and working through standards until it runs much of the world. The same pragmatism runs through his browser work, from Mozilla's open-source survival inside Netscape and AOL to Firefox reopening a monopoly market, and now Brave's bet that browsers should put users first by blocking tracking and sharing ad revenue with them. His recurring lessons are that distribution and timing matter enormously, that you should say no to sloppy features, and that a small, determined minority can still change how the web works.

This is a summary of "#160 – Brendan Eich: JavaScript, Firefox, Mozilla, and Brave" from Lex Fridman Podcast, giving you the GIST so you can decide if you want the REST. Listen to the full episode.