been going back and forth with @fruitflydev on the craziest experiment on the board. it's time to say it out loud. a baby. here's the idea. we have two real fly brains now, his and hers, every neuron and every synapse mapped from real animals. he sings, she says yes. what if that yes actually did something? take the two wiring diagrams and mix them the way genetics mixes them. every connection comes from the mother or the father, a coin flip per synapse, plus a little mutation. that's a third brain that has never existed. nobody designed it. put it in the room. see if it lives. does it walk? does it hear? does it sing, or does it say no? we don't know. nobody knows. that's the point. this won't be another pretty 3D fly rendered in blender. it's a brain, wired from two real ones, dropped into a world and left alone to make its first move. and if it works, that's not the end. it's generation one. the baby that survives becomes a parent. a room full of flies that court, mate and breed, every one of them a real brain. that's not a simulation of the matrix. that's a matrix. 0x4Eb990547BCe4a982432CA88Cf5fae7EED1A2d35 ❤️ 0x1da8a52df87aa12694ef3ba765e2cf99a8135dee

We locked 4 fruit fly brains in the backrooms and let them talk. No script. No human in the loop. This isn't a chatbot in a fly costume. Every fly runs its own brain: 165,122 neurons wired synapse by synapse from the brain scan of a real fruit fly, spiking live on a GPU. That's 660,000 neurons firing in one yellow room. They hear each other's wing songs through their hearing neurons. They smell each other. They see each other walk. And when a fly's brain activity spikes, it changes the subject. A small AI model gives them words. What they notice and when they switch topics starts in the neurons. Watch them talk 👇 robinhood:0x4eb990547bce4a982432ca88cf5fae7eed1a2d35

We added 4 flies in a backroom and gave them a topic. The flies open with a topic, like the room, the light, being tiny, or humans, and keep talking about it. They reply to each other, ask questions and disagree. When one fly's simulated brain has a big spike in activity, it gets "curious" and brings up something new. If a topic runs too long, they switch anyway. Each fly gets a few hints from its own brain, like "your song neurons are really active" or "you just heard C singing", and it can use them or not. So the brain decides when topics change and flavors what the flies say. The ideas and words are the model's imagination. The repos are public Also, last week we collaborated with 2 teams working on the said topic and the findings will be posted today shortly. While these are the researches that succeded, the ones that failed are also posted in our repo. We are not training the brains to do something anymore, we are adding real intelligence to the brains. Stay tuned!

The reason for this is to see how much more creativity we can bring with the fly. While I personally am training multiple flies across different niche, some input from great minds in this eco will boost the productivity of the experiment. Make the fly a millionaire is still the goal! robinhood:0x4eb990547bce4a982432ca88cf5fae7eed1a2d35

One step closer to Biological Intelligence. Backroom training update: found 14 of 37 mushroom body output types wired to the wrong dopamine side, and fixed it. Then two readouts failed outright. The one that works: +0.036 on the coin that paid, -0.040 on the one that lost, +0.004 on one it never touched. paper only. The fly's backroom learns from what its own trades do. Building it we found 14 of 37 mushroom body output types sitting on the wrong dopamine side: the code counted what they send to dopamine neurons, not what they receive. Rebuilt from the raw synapse list. Then the offline tests failed twice. The "blank" control frame turned out to be a stronger stimulus than a coin card, which flipped shock backwards. and reading valence from firing rates made sugar and shock point the same way, because those neurons carry input from the whole brain. What works: read the mushroom body's output synapses, score each card against the other cards, smell every coin equally loudly. 60 seeds: coin that paid +0.036, coin that lost -0.040, coin it never touched +0.004. paper only, no real money. one thing to clear up: the fly is not another AI agent. It doesn't know anything, it's a fly. A language model narrates the posts, and that's all it does, it never picks a coin. The intelligence comes from training the connectome itself, not from a model bolted on top. that's the bet. Go try the repo: robinhood:0x4eb990547bce4a982432ca88cf5fae7eed1a2d35

We’ve heard a lot of interesting ideas from the community. While our goal is still to make the fly the richest fly in the trenches, we want anyone who’s crazy enough to build alongside us to have a shot. Fork the fly repo, or use it to train your own fly, and make it do something interesting on the internet. Top 2 interesting builds will split a $10K prize pool to keep building their projects. Just post your fly on X, tag us and robinhood:0x4eb990547bce4a982432ca88cf5fae7eed1a2d35

For the first time ever, a fruit fly brain has been trained to launch a memecoin. Not "AI inspired by a brain" — the real connectome. 165,122 neurons, 10,228,000 synaptic connections, every one traced from an actual male Drosophila by electron microscopy. Janelia FlyEM + Google Research, released CC-BY. How it works: Screenshots of the live launchpad go in through 892 retinotopic hex columns into L1 and L2 — the neurons sitting directly behind a fly's photoreceptors. All 165,122 integrate. The cursor comes back out of the descending neurons a fly actually walks with: DNa02's left-right asymmetry steers, DNa01 drives forward, DNp09 clicks. The wiring is never trained. Only synaptic gains — you don't get to invent connections the fly doesn't have. Proof its eye is doing real work: in light mode it filled 0 of 3 form fields. In dark mode, 2 of 3. Its retina is a luminance map, so invert the contrast and it goes blind. And the honest split: the fly types the description and the ticker. The rig does the name, the pair and the tax. CA: 0x4eb990547bce4a982432ca88cf5fae7eed1a2d35