{"id":38,"date":"2026-09-29T23:40:36","date_gmt":"2026-09-29T23:40:36","guid":{"rendered":"https:\/\/technewshome.com\/index.php\/2026\/09\/29\/google-fruit-fly-brain-doom\/"},"modified":"2026-09-29T23:40:36","modified_gmt":"2026-09-29T23:40:36","slug":"google-fruit-fly-brain-doom","status":"publish","type":"post","link":"https:\/\/technewshome.com\/index.php\/2026\/09\/29\/google-fruit-fly-brain-doom\/","title":{"rendered":"Google mapped a fruit fly brain, so engineers immediately made it play Doom 6,000 times"},"content":{"rendered":"<p class=\"lede\">Last week, neuroscience achieved a genuine milestone: the complete map of an adult fruit fly&#8217;s brain and central nervous system. This week, that brain is playing Doom. Badly. And the gap between those two sentences tells you everything about how science actually moves forward.<\/p>\n<aside class=\"key-takeaways\">\n<p class=\"key-takeaways__title\">Key Takeaways<\/p>\n<ul>\n<li>Google Research, HHMI Janelia and the wider community released MaleCNS v1.0, the first complete connectome of an adult male fruit fly.<\/li>\n<li>AI stitched millions of 2D images into a 3D model reconstructing more than 166,000 neurons.<\/li>\n<li>Coinbase engineer Alex Wormuth connected the simulated brain to the original Doom within days.<\/li>\n<li>Each game frame stimulates sensory neurons; neural activity maps to controls; damage triggers two PPL101 dopamine cells as reinforcement.<\/li>\n<li>The fly has passed 6,385 attempts and remains, charitably, a work in progress.<\/li>\n<\/ul>\n<\/aside>\n<h2>The milestone underneath the meme<\/h2>\n<p>Strip away the Doom overlay and MaleCNS v1.0 is serious science. A connectome is a wiring diagram of a nervous system: every neuron, every synapse, every connection. Fruit flies have been a workhorse of biology for over a century because their brains are complex enough to generate real behavior yet small enough to map completely. Completing the adult male connectome, and releasing it publicly, gives researchers a full reference circuit for studying how brains produce actions.<\/p>\n<p>The technical lift was enormous. The team used AI to stitch millions of two dimensional microscopy images into a coherent three dimensional reconstruction of more than 166,000 neurons. What once took human tracers years per region now takes machines days. This is the same pattern we see across research, where AI accelerates the boring parts so humans can argue about the interesting ones.<\/p>\n<h2>So why Doom?<\/h2>\n<p>Because a wiring diagram is not a working brain. Knowing the connections does not tell you what the circuit does when it receives input. Engineers like Alex Wormuth solve this by embodiment: connect the simulated connectome to an environment, feed it sensory data, and see what its activity produces. Doom is perfect for this. It renders simple 3D spaces, exposes a tiny set of controls, and provides clear feedback in the form of damage.<\/p>\n<p>The setup is elegantly direct. Each Doom frame stimulates the fly&#8217;s simulated sensory neurons. The resulting neural activity is decoded into game controls. When the fly takes damage, two specific dopamine neurons, labeled PPL101, receive a stimulus, providing a reinforcement signal analogous to how real flies learn to avoid harm. A live feed shows both the in game view and a third person render of the fly&#8217;s digital body wandering the corridors.<\/p>\n<h2>Six thousand failures are the point<\/h2>\n<p>At the latest count, the fly had attempted the game 6,385 times without anything resembling competence. It is tempting to laugh, but every failed run is data. The experiment tests whether a raw connectome, plus a simple reinforcement signal, can bootstrap behavior, a question with implications for neuroscience, robotics and AI safety research. If structured behavior emerges from wiring alone, brains are more deterministic than we assumed. If it does not, learning rules and body feedback matter more than the map.<\/p>\n<p>Either result teaches us something, which is more than can be said for most gaming marathons. It also echoes debates in the AI world about emergent behavior, the same theme behind the <a href=\"https:\/\/technewshome.com\/anthropic-ceo-ai-internet-takeover\/\">Anthropic CEO&#8217;s warning about agent swarms<\/a>: systems doing things nobody explicitly taught them.<\/p>\n<h2>What comes next<\/h2>\n<p>Expect follow up experiments with richer environments, better reinforcement schedules and comparisons against the female connectome. Longer term, whole brain emulation research, from worms to flies and eventually beyond, depends on exactly this kind of playful, rigorous testing. Today&#8217;s Doom playing fly is tomorrow&#8217;s validated simulation method.<\/p>\n<p>For a very different kind of long term engineering, read our report on <a href=\"https:\/\/technewshome.com\/curiosity-rover-wheels-14-years-mars\/\">what 14 years on Mars did to the Curiosity rover&#8217;s wheels<\/a>.<\/p>\n<h2>From worms to flies: a short history of connectomes<\/h2>\n<p>The field started, famously, with a worm. The nematode C. elegans, with its 302 neurons, had its nervous system fully mapped decades ago, and that map has anchored generations of research into how circuits produce behavior. The fruit fly connectome is a different league: hundreds of thousands of neurons, rich sensory systems, learning, memory and navigation. Each step up in organism complexity has historically unlocked a matching step in understanding, which is why the MaleCNS release landed with such force in the research community.<\/p>\n<h2>What researchers will do with the map<\/h2>\n<p>Practical applications are already queueing up. Neuroscientists can trace the circuits behind specific behaviors, from courtship to navigation, against a complete reference. Drug researchers get a testbed for understanding how compounds alter circuit dynamics. And the AI community gets something it has wanted for years: a complete biological network to benchmark against artificial ones. The Doom experiment is the playful tip of a very serious iceberg, and the fact that the data is public means the next surprising result could come from a university lab, a company, or a particularly determined hobbyist.<\/p>\n<h2>The engineering behind the experiment<\/h2>\n<p>What makes the Doom setup more than a stunt is its faithfulness to biology. The simulation does not hand the fly a game controller; it renders Doom&#8217;s frames into patterns of sensory neuron stimulation, then reads the connectome&#8217;s spontaneous activity and decodes it into movement. The reinforcement channel, those two PPL101 dopamine neurons, mirrors the real circuit flies use to associate experiences with harm. Even the failure is informative: if a complete wiring diagram plus a biologically grounded learning signal still cannot navigate E1M1 after 6,385 tries, then instinct alone is not enough, and something about real bodies, real chemistry or real development matters more than the static map suggests. That is a genuine scientific result, delivered by a virtual insect repeatedly walking into walls.<\/p>\n<p>And yes, we will keep watching the counter. Somewhere between attempt 6,385 and whatever comes next, the question stops being &#8220;can a fly play Doom&#8221; and becomes &#8220;what does it take for a mapped brain to learn anything at all&#8221;. The answer, whenever it arrives, will matter far beyond gaming.<\/p>\n<p class=\"source-note\">Source: <a href=\"https:\/\/research.google\/\" target=\"_blank\" rel=\"noopener nofollow\">Google Research<\/a> and HHMI Janelia Research Campus, MaleCNS v1.0 release (September 2026).<\/p>\n<p class=\"article-cta\"><em>Following connectome research? Share papers or questions with the newsroom via our <a href=\"https:\/\/technewshome.com\/index.php\/contact\/\">Contact page<\/a>.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Days after Google and HHMI Janelia released the first complete connectome of an adult fruit fly, an engineer wired it into Doom. 6,385 attempts later, the fly still has not beaten it.<\/p>\n","protected":false},"author":2,"featured_media":39,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5],"tags":[],"class_list":["post-38","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-science"],"_links":{"self":[{"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/posts\/38","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/comments?post=38"}],"version-history":[{"count":0,"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/posts\/38\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/media\/39"}],"wp:attachment":[{"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/media?parent=38"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/categories?post=38"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/technewshome.com\/index.php\/wp-json\/wp\/v2\/tags?post=38"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}