A 3D fruit fly on macOS desktop powered by the real FlyWire connectome

Project impressions & demos

• Many find the desktop fruit fly simulation “cool” and “astounding,” with several running multiple flies and praising it as the kind of hack they come to HN for.
• Some want a simple demo video or GIF; they argue the author could have saved collective time and compute by including a 10‑second screen capture in the repo.
• Others note it’s trivial to clone/build on macOS but point out that non‑Mac users can’t run it at all.

What the connectome simulation is (and isn’t)

• Commenters appreciate the open-source transparency compared to hype‑driven startups.
• Several stress that calling this “controlled by the connectome” is somewhat misleading: scripted behaviors are mapped to neural triggers, so the connectome may function more as a structured trigger/entropy source than a full behavioral model.
• Key limitation: only connectivity is known; synaptic weights and many biophysical details are missing.
• Some note the code uses spiking‑neuron style activation and ask whether synaptic weights can change dynamically, seeing this as critical for realism.
• Others point out broader biological gaps: real neurons use spike timing; hormones and other modulators are not modeled; butterflies and worms are mentioned to show that connectomes alone may be insufficient (though this is debated).

AI‑generated README and style tics

• Multiple comments dissect the README as clearly LLM‑written, criticizing phrases like “honesty section” and characteristic hyphenation and tone.
• There’s frustration that AI docs feel like self‑advertising and rely on weasel words, even as some concede the terminology (e.g., “integrate‑and‑fire”) is standard in neuroscience.
• Discussion touches on system prompts trying (and failing) to suppress certain stylistic tics.

Ethics of simulating brains

• A long subthread debates whether simulating small nervous systems can be unethical.
• One camp: it’s “just electrons,” with no moral status beyond a normal program.
• Another camp: if simulations can in principle realize consciousness or suffering, we should treat them cautiously; even low probability of “digital suffering” may matter.
• Broader arguments arise about who deserves moral consideration: fruit flies, bacteria, lice, cows, and plants, with differing views on how far empathy should extend.

Platform choices and portability

• Some criticize making it macOS‑only in 2020s; they expect at least Linux/Windows support or a web version.
• Others defend targeting the author’s own OS and avoiding cross‑platform compromises.
• There is active work (by commenters) on porting the simulation to run in a browser, possibly with multiple flies, interaction via camera, and 3D rendering.

Broader implications & related work

• Commenters connect this to other projects like OpenWorm and fruit fly body simulators, suggesting integration for full embodied behavior.
• Thought experiments appear about:
  • Simulating organisms from DNA as a “ROM” inside a detailed physics sandbox.
  • Whether full human‑brain connectomes could yield “digital humans.”
• Some are optimistic that large neural simulations plus physics could eventually approximate biology; others doubt that matrix multiplications can capture the full chemical complexity.
• Fictional works about uploaded minds and digital suffering (short stories, TV series, novels) are cited to frame the ethical and philosophical stakes.