recently, the most advanced fly brain connectome dataset until now was released as result of the years-long project by collaboration of google research and hhmi janelia. the dataset consists of the whole brain and nerve cord of a healthy adult male drosophila, 11,700 types of neurons and neural pathways&synaptic data.
developers seeing this started to simulate the fly brain architecture on computer to train reinforcement-based models to achieve specific tasks; such as playing games, solving puzzles or doing actions. the system is simple: fly brain model is taken, fly’s neural pathways are linked with results of specific functions (e.g dopamine pathway is triggered when the fly captures the health orb in the game), and reinforcement learning steps are executed in order to “teach” the fly to “live” in the simulation setup that was built. people have been building very creative setups: from flies that parallel park to flies that writes computer programs, i’ve seen many interesting projects (i had built a doom simulation using older fafb v768 dataset by princeton before, although i used a frozen connectome and optimized doom in order to preserve real fly behavior inside doom instead of training a model on fly brain architecture to play doom).
in addition to the developers building those simulations, i’ve seen many non-developers arguing about the ethics of such experiments. the common point/argument was something close to “the connectome has a fly’s nervous system running with a ‘synthetic’ environment it interacts with; so how’s it different from using a real animal for those experiments?”. if we see the fly’s brain as “a carrier of a conscious mind”, the argument becomes acceptable on the grounds that a perfectly copied digital remake of the same structure would also carry such mind.
the question before this, however, would be “is the fly a conscious organism in the first place?”. in the scientific consensus, the answer would be yes; considering the 2012 Cambridge Declaration on Consciousness and more recent 2024 New York Declaration on Animal Consciousness. therefore, a real fly is a conscious organism, and if a simulated environment “perfectly” copies its brain, that copy could also be considered “conscious” & should be examined within the same standards.
even though it might make me “soulless”, reckless or rude; i’ll say that nobody gives a fuck about a fruit fly. sorry vegans. i dont.
the bigger, and much more concerning question here will open when (or if) somebody manages to model the human brain: what level of humanism we’ll approach with to a simulated human brain? and what separates a real human with its thoughts, feelings and rights from a simulated one? after all, the conscious human mind also consists of an extremely complex and fascinating biological network and it interacts with an external environment; not different from a potential synthetic brain model interacting with a synthetic environment perfectly copied. have fun.