Last week, X was irked by some posts about a so-called virtual “absorbed fly”, which was promoted by AI hype accounts and excited commenters who didn’t understand what they were excited about.
The videos come from San Francisco-based Eon Systems, which says it’s working toward “digital human intelligence” and claims it wants to create a full digital simulation of a mouse brain within the next two years — a timeline, to put it generously, that’s ambitious. Co-Founder Alexander Wiesner-Gross shared The original clip was publicly released, calling it “the world’s first incarnation of a whole brain simulation that generates multiple behaviors” and hinting at an impending technological singularity. CEO Michael Andreg Posted a different cut, describe it As a “real uploaded animal”.
And that was it for the proof: no detailed methods, no scientific papers, no independent verification, just videos of a digital fly hopping, eating, and rubbing its legs together.
We have uploaded a fruit fly. we took @FlyWireNews The connectome of the fruit fly brain implemented a simple neuron model (@philip_shiu Nature 2024) and used it to control a MUSCO physics-simulated body, thereby closing the loop from neural activation to action.
Some things I want… pic.twitter.com/Qnlu3INs33
– Michael Andregg (@michaelandregg) 8 March 2026
related to AI accounting book But x And reddit amplified The clip and caption were repeated as fact. Anticipated drive-by support from people like Elon Musk (“Very good”), brian johnson (“It’s amazing”), and peter diamandis (“It’s a living being…online”) added fuel to the fire. Then content farms came along and repackaged the whole thing as “news.” are celebrating first time brain upload And I am asking “Are humans next?(yes, those too referred to math questionAnd spoiler: we’re not next).
“This is, in our view, a real uploaded beast.”
The internet was buzzing. The proof was still two short videos on X. If you’re going to tell the world that you just achieved the most important scientific milestone in human history, you better bring receipts.
Andregue attempted to provide some clarity on Thread It was somewhat of a grab bag of warnings, somewhat vaguely described scientific terms, and somewhat solid-sounding numbers like “91% transaction accuracy” I sat with that metric for a while, and I still don’t really know what it means, and I spent a large part of my master’s degree studying animal behavior. Still, he Insisted That “this is, in our view, a real uploaded animal.”
I sent Andrag a message on LinkedIn asking for more details. He replied with a link to a recent blog post written by Eon published Titled “How the Aeon Team Built a Virtual Embodied Fly.” It wasn’t a scientific paper, but I guess it was something.
for experts The Verge That blog didn’t go nearly far enough, but it outlined a much more cautious line than the post on X, so much so that it doesn’t say “it’s a real fly.” Professor Shahab Bakhtiari, who leads the Systems Neuroscience and AI Lab at the University of Montreal, said the initial post had “obscured important details” about the work, but the new blog provided more context. “But it came a little late, and is insufficient to fully validate the claims,” he said. They will be expected to have a detailed technical report that will include details of things like software, code, and simulation environments that will allow other scientists to reproduce and evaluate the work.
Alexander Bates, a research fellow in neurobiology at Harvard Medical School who studies fly brains, echoed Bakhtiari. He said the group “under-delivered” and while the blog provided more details about what the team did – they pieced together existing large-scale projects such as a detailed map of a fly’s brain, a physical simulation of a fly’s body, and models simulating how these interact in a virtual environment – ”For a claim of this magnitude, I would expect something that uniquely conveys the entire approach.”
Bates also said that the virtual fly’s behavior should be evaluated based on real data and “clearly defined metrics”, adding that the 91 percent figure was still not explained in the blog post. “Besides, flies don’t fly.”
Bates said The Verge He understands that “strong framing and publicity can matter for fundraising,” but insists that Eon’s claim of “real uploaded animals” is not credible. Aran Naybi, a professor of machine learning at Carnegie Mellon University, said the group was “not even close” to capturing the fly’s entire brain, showing connections between cells, but not important details like neurotransmitters or how strong the connections are between different nerve cells. The motor system is also not “true upload”, he said. “We’re honestly not even simulating its brain in silico“
Well, let’s say Aeon actually did do it. It exactly copied the fly’s brain. The whole thing. Every last part. Do we have a digital fly now?
Yes. no maybe. I don’t know, and maybe you don’t know either. Not even an eon. The blog shed light on the weighty definitional questions at the core of the easy upload claim: What, exactly, counts as a fly? When we think of flies, we don’t think of a ton of behaviors or neural connections. We think of a fly. Is it enough to reproduce some fly-like behaviors in simulation? Does a fully mapped brain count in a virtual VATS? Or does “flying” mean the entire, disorganized biological package – a body, cells, metabolism, and whatever else is considered “memory” or learned experience during its life?
And this is the simple version of the problem. The thing on the screen is clearly not a fly at all. It is a combination of neural wiring, programming and other information from many different animals linked together. This is useful when modeling, but in this case what organisms can we actually make a meaningful claim about being uploaded? Technically speaking, it’s a copy, not an upload, which comes with obvious and profound implications the hype easily overshadows: you can make two, or 10, or 10,000 “identical” flies. so what?
Normally, I wouldn’t expect a startup to solve any major metaphysical problem – philosophers have been debating this for centuries – but they’re the ones saying they’ve found the “real uploaded beast.”
Even the experts I spoke to weren’t convinced the term made sense. Bakhtiari said it is still an “open question” whether a “real uploaded animal” is even possible. Philosopher Jonathan Birch of the London School of Economics put it bluntly: “I don’t think we should ever say ‘uploaded animal,'” he said. Eon’s goal is “whole-brain simulation,” he said, leaving the rest of the animals behind.
“…this fly is conscious in a limited sense; it can smell, see, taste, etc.”
Philosopher Tom McClelland of the University of Cambridge said that biology is important for behavior. “So the best thing they can do is upload part of the fly’s brain and thus upload part of the fly.”
After spending some time following his viral post, I asked Andreg if he stood by his claim. “Yes,” he told me. In fact, he added: “We (the research group and its academic collaborators) think that this fly is conscious in a limited sense; it can smell, see, taste, etc.” (I won’t go into the whole consciousness thing.) He described the system as a kind of “MVP”, a minimal viable product of an uploaded beast, with “a lot of limitations”. I don’t quite know how to depict a minimum viable fly. It’s a fly, not an app. MVP is a matter of tech startup, not science.
When I went to Andreg a second time – this time after talking to experts and criticizing them – he still stood by the original claim, but with even more caveats. He acknowledged that the piece “is not a perfect replica of a fly,” and said that Eon had never said it was so. “I don’t think of uploading as a binary concept,” he told me, describing “different levels” of uploading and acknowledging that we don’t yet know how much biology is needed to obtain important information. “There is still a lot of work to be done to achieve the level of uploads we want for ourselves someday.”
