Jonathan Blow

• 1. Introduction

  ›

  Jonathan Blow, an American game designer and programmer, once summed up his working philosophy in one blunt line: "The job isn't to be in a community; the job is to make a good game." He said it after some of his comments about independent development drew social-media attention he found troubling. The sentence fits a career that has run against the grain of the games business. What drives a person to spend years on a single puzzle game? Why would he insist on paying for it himself? And what does he believe is wrong with the way software gets made?

  715663

• 2. Early Machines

  ›

  Jonathan Blow was born in Southern California on the 3rd of November 1971, the fourth of five siblings. His father worked as a defense contractor for TRW, and his mother was an ex-nun. The family was Christian and attended church regularly.

  A fifth-or-sixth-grade computer class in Northern San Diego County put a VIC-20 in front of Blow, and that machine gave him his first taste of programming. When his parents noticed his interest, they bought him a TRS-80 Color Computer. On it he learned BASIC, often working from RadioShack exercise books. By high school he was writing games on a Commodore 64, and some of them drew on Indiana Jones and Pac-Man.

  As a player, Blow gravitated toward the text adventures made by Infocom, and his favorite was Trinity. He said it shaped how he thinks about games in a relatively deep way. In college his favorite game was Netrek, and early in his career he often played Counter-Strike.

  In 1989 Blow enrolled at UC Berkeley as a physics major, then switched to computer science because he "just felt called in that direction". He double-majored with English. He was a member and president of the Computer Science Undergraduate Association, and he belonged to the eXperimental Computing Facility, an undergraduate computer-interest organization. He also wrote science fiction, which he published under a pseudonym.

  After five years at Berkeley, Blow dropped out with less than one semester to go. He later explained: "I was really depressed about being at school, I didn't like it. I didn't have a good time." The eXperimental Computing Facility had still paid off, though, because it was where he met a fellow member named Bernt Habermeier.

  4597985

• 3. Boom, Bust, Contract Work

  ›

  Six months at a "really boring" enterprise software company was all Blow could take before he moved to a contracting role at Silicon Graphics. There he ported Doom and Doom II to a set-top box.

  In early 1996, Blow and Bernt Habermeier founded Bolt-Action Software in Oakland, California. An artist joined later, and the team built Wulfram, a 3D action-strategy game for up to 32 players. Each player took control of a heavily armed hovertank. At its height the company employed 14 people.

  The dot-com crash folded Bolt-Action Software in 2000. Looking back in a 2020 interview, Blow said 1996 was the hardest time in history to start a game company, because the industry was shifting from 2D to 3D titles. He took a hard lesson from the failure: "we went broke, and I was burned out for several years after that from working hard... but that's how I became a good programmer".

  After closing his first studio, Blow worked as a contractor for game studios with large budgets. His credits included Oddworld: Munch's Oddysee, Deus Ex: Invisible War and Thief: Deadly Shadows. In 2002 he co-founded the Experimental Gameplay Workshop at the Game Developers Conference, a showcase for prototypes with new mechanics, new genres or new mediums. Around then he also wrote The Inner Product, a monthly technical column for Game Developer magazine.

  A move to New York City brought Blow into an IBM research project about servers built on cell processors. He pitched IBM a proof of concept: a physics-intensive, online, multiplayer game about giant robots attacking a town. Blow and Atman Binstock did most of the programming. Blow wrote the client-side code, the graphics and the gameplay, while Binstock wrote the server's physics engine from scratch. After the final report went to IBM, the team took the game to Electronic Arts, who Blow said were not impressed.

  Other contract jobs included particle effects for Flow on the Sony PlayStation 3, code review after MTV bought Harmonix, and programming on the iPod music game Phase. Blow described the era this way: "I was just stumbling forward like people do sometimes, and doing the best that I knew how to do, which at that time was programming." In December 2004 he finished a prototype for a 2D puzzle-platform game about manipulating time.

  578910

• 4. Braid

  ›

  By Blow's own estimate, he spent more than $180,000 of his own money developing Braid. He began turning the prototype into a real game five months after he made it, and by December 2005 the first version was complete. Much of the work happened part-time, because he also did consulting and martial arts training.

  Blow felt the first version's graphics and style "looked extremely amateur", so he hired David Hellman to create all of the game's art. For the story he drew on favorite books and films, including Invisible Cities and Mulholland Drive. The narrative arrives through text between worlds, environmental art and the gameplay itself. Players have interpreted it in many different ways.

  Kung fu training began for Blow at the same time as Braid, and near the end of development he was doing 15 hours a week. The practice taught him meditation techniques that helped his work. He said: "I don't know if I would have finished Braid if I wasn't doing kung fu."

  In mid-2007 Blow signed with Microsoft to release Braid on its XBLA distribution platform. He felt the time spent on the XBLA certification process would have been better spent polishing the game. Even so, he called Microsoft "very hands-off" about game design and said that "the final game is exactly what I wanted to put there".

  Braid was released in August 2008 to universal acclaim. It was an immediate sensation, it made Blow a millionaire, and it was one of the earliest indie games on seventh-generation consoles. A GQ retrospective noted that the market was then dominated by violent, repetitive first-person shooters, yet Braid made an earnest effort at an artistic statement. Its puzzles were described as "tough" and "formidable".

  In 2010 Blow co-founded the funding organization Indie Fund with several other successful independent developers. Two years later he appeared in the documentary Indie Game: The Movie, where he discusses developing and releasing Braid. In 2014 he stated that Braid's sales had earned more than $4 million.

  215162021

• 5. The Witness

  ›

  Nobody ever hands a player of The Witness a written rule. The first-person game asks players to solve puzzles by drawing paths, and the puzzles themselves teach the rules without words. Blow wanted to build a game around non-verbal communication. He hoped players would experience "miniature epiphanies over and over again". He estimated that solving every puzzle would take more than 80 hours.

  Work began in 2008. Blow prototyped several game ideas before choosing the one he liked best, even though it was a 3D game he "absolutely didn't want to do". As development went on, he hired developers full-time and founded a company called Thekla, Inc., of which he is president. The game was revealed to the public in 2010, when three people were working on it full-time.

  Dancing kept Blow going during these years. He practiced tai chi early in development, and he went out dancing several nights a week while the game took shape. Dancing helps him generate ideas. In the 1990s he even carried a notebook to clubs to write down programming ideas that arrived on the dance floor.

  Blow had hoped to ship The Witness in 2013 as a launch title for the Sony PlayStation 4, but the goal slipped as the game's scope grew. By 2015 the core team had grown to eight people. At that time it was very rare for a small independent studio to spend more than seven years on a game.

  The Witness arrived on Windows and the PlayStation 4 in January 2016. Blow reported first-week revenue of more than US$5 million, and the game also ranked among the top downloads on illegal BitTorrent websites. It drew critical acclaim, several BAFTA and Game Developers Choice Awards nominations, and places on best-of-the-decade features from IGN, Polygon, NME, CNET and National Post. Time called it "categorically defiant" and said its reception was "widely quantified as a game created by a genius for geniuses".

  Years earlier, while the game was still unfinished, Blow had given a talk on his Twitch channel in September 2014. Its subject was a possible new programming language for game development.

  123282955

• 6. Freedom and Cost

  ›

  Close to six million dollars went into The Witness, and Braid's profits covered most of it. Braid itself had a relatively large budget for an independent game of its day. The development times of both games were very rare among independent studios at the time.

  Both games run on custom engines. Blow said that when The Witness began, off-the-shelf engines such as Unity and Unreal Engine were not capable enough, and building his own took a large amount of development time. He also wants games that stand the test of time. Owning the engine means he can keep supplying the game to people, which off-the-shelf engines do not guarantee.

  Thekla exists because Blow wanted creative freedom, not because he aspired to run a company. By funding The Witness almost entirely from Braid's profits, he avoided bending the game to the whims of a publisher. The team on The Witness said he gave them plenty of autonomy. As for schedules, Blow put it plainly: "we don't do it like Electronic Arts. We don't pick a quarter and ship the game. We're just trying to make the best game."

  Braid artist David Hellman saw the same attitude up close. He said Blow was not stressed about time or budget, and that for him "the game always came first".

  Blow says he tries to make something he would want to play. He wants to understand the world from many perspectives and to uncover truths about the universe through design. He enjoys exploring the permutations of a simple idea. His ideal player is someone who "is inquisitive and likes to be treated as an intelligent person". He hopes the design skills built on such complicated games have made his programming language better designed than most.

  22607250

• 7. The Jai Language

  ›

  Go and Rust were both on Blow's list when he went looking for an alternative to C++, the language Thekla used for The Witness. Neither satisfied him, and he decided he wanted to create a language of his own. C++ is powerful in some ways, he said, "but it makes software development a lot harder than it should be". In 2020 he described modern C++ as a terrible language.

  Blow estimated that a language designed for game development could cut typical development time by at least 20% and make programming more enjoyable. He also judged it relatively easy to create compared with a game like The Witness. He hopes it will let programmers build more functionality with less code, and he was partly motivated by improving programmers' quality of life.

  In 2014 Blow began designing the language, codenamed Jai, and started writing its compiler. The first year and a half was part-time, because Thekla was still shipping The Witness. In mid-2016 he began full-time work on the language, a game engine written in it, and a new puzzle game called Order of the Sinking Star. In 2023 he called it "the biggest single-player puzzle game that anybody's ever made". He estimated 250 hours at a normal pace and more than 500 hours for completionists.

  Building the language, the engine and the game together lets Blow test the language's design and adjust it early in its life. As of 2023, the Jai compiler was in a beta release.

  In August 2020, Thekla announced Braid, Anniversary Edition, a remaster. Blow promised it would stay faithful to the original, with no "Greedo shoots first" changes. It was to include more than 15 hours of developer's commentary. Planned for early 2021, it slipped, and it was eventually released on the 14th of May 2024. Blow was frustrated by the weak reception. He said Thekla could not afford to pay any of its employees to work on the Jai compiler.

  Back in 2013, Blow had started a prototype for a single-player game that was not a puzzle game. By 2018 he said it held 40-50 hours of playable content. He intends to build it on the engine being developed for Order of the Sinking Star once that engine matures. He plans to work on it for 20 years, releasing it in installments that each make the game larger and more complex, partly to expand his own design abilities.

  3940414885

• 8. Prickly Critic

  ›

  VentureBeat credited Blow with a "reputation for doing outstanding work". Others have called him a "prickly genius" and the industry's "most cerebral developer, but also as its most incisive and polarizing internal critic". Blow says: "I honestly say what I think about games, and I honestly say if I think something is good or not, and why." Stephen Totilo of Kotaku observed that the criticism targets industry trends, not individuals or specific games.

  Storytelling is one target. Blow thinks mood, character and setting can work well in games, but he considers games a terrible medium for storytelling overall. As of 2016 he rated the storytelling in games as significantly lower in quality than in literature. In 2018 he argued that many contemporary games contain self-sabotaging designs that weaken their own structure. Because such designs already existed nearly a decade earlier, he saw stalled progress in the medium.

  Money models are another. Blow does not consider microtransactions inherently unethical, but he thinks many phone games are "just pretending to be games in order to have a microtransaction button or show ads". He considers social network games evil. SimCity and FarmVille look similar on the surface, he noted, but SimCity is a creative, problem-solving activity while FarmVille exists to hold a player's attention as long as possible. He has also warned that massively multiplayer online games, which hook players with fake rewards, may cause real harm.

  Stephen's Sausage Roll, which Blow said in 2017 "may be the best puzzle of all time", became his example of a stagnant scene. He noted that it was rarely discussed in development circles. In 2019 he said releasing a game was easier than ever, but that progress in game design "has not been as large as people assume". He is uncomfortable being called an indie developer, because he feels that scene has changed into something he does not belong to.

  Software as a whole draws his scorn. In a 2020 interview he said: "I think we're now in a situation where everybody in the world is flooded with low quality software, and everybody wishes that they had higher quality software." In 2021 he described much of what programmers do as dealing with unnecessary complexity.

  In December 2025, game developer Alan Hazelden, who had licensed two of his own games for use in Order of the Sinking Star, made a statement against Blow over his political views. He cited Blow's support for Donald Trump, negative views on diversity, equity and inclusion, and opinions on COVID-19 vaccinations. Two other developers from the same project, Sean Barrett and Jonah Ostroff, echoed the statement on social media. Blow, described as intensely private, remains active on X, YouTube and Twitch, where commentary on game development sits alongside personal political opinions.

  5663707576