Silicon Valley startup Eon Systems claims to have successfully uploaded a fly's "mind" and placed it in a simulated environment. The uploaded "mind" can control a digital body and respond to sensory inputs with natural behavioral reactions, such as walking, grooming, and foraging, with an accuracy rate of up to 91%. 
Dr. Alex Wissner-Gross, co-founder of Eon Systems, stated, "This is not animation, nor is it a reinforcement learning strategy that mimics biology. This is a copy of a biological brain, connected neuron by neuron via electron microscopy data, running in a simulated environment, driving body movement." 
Although there are currently no new scientific papers to support this claim, Eon Systems has made the relevant code repository and a demonstration video publicly available on GitHub. The research is based on a computational model of the fly brain published in 2024 in the journal Nature by Eon senior scientist Philip Shiu, which contains 125,000 neurons and is known as a connectome. For this upload experiment, they did not scan the fly's body but instead provided the "mind" with an off-the-shelf version of NeuroMechFly v2 for its use. However, the "mind" is currently unable to form memories. 
Eon's claim about uploading a "mind" has sparked some controversy. Some argue that it is more accurate to describe it as a "copy," while others believe it is merely a model. Tech entrepreneur Chomba Bupe holds the latter view, arguing that it is just a simplified model that can predict some neural activity in the fly brain. "Mind uploading means capturing all cognitive aspects of a being, including its consciousness, into a computer," he commented. "Using a simplified computational model of an organism is not uploading, but basic simulation." However, he also acknowledged that the work is still "great" and worth exploring further. 
After perfecting the fly brain, Eon Systems' next goal is to upload the brain of a mouse, which has 560 times more neurons than a fly. The ultimate goal is to upload a complete human brain.
Another eerie scientific breakthrough is that a developer, using the Cortical Labs API and programming in Python, has successfully taught a clump of living human brain cells how to play the classic first-person shooter game Doom. 
This biocomputer/wetware contains 200,000 neurons and, while it doesn't play Doom at a high level, it's better than just shooting randomly.