Ariel Zeleznikow-Johnston

Startup survival requires investment, funding follows excitement, and excitement follows headlines, not careful caveats.

This bold approach may even feel obligatory when an organization's stated mission is solving brain emulation as an engineering sprint, not a decades-long research program.

But the history of science, and the gap between what Eon demonstrated and what uploading actually requires, suggests that there is likely no shortcut through the long slog ahead.

Because in all probability, before anyone can truthfully claim to have uploaded a fly, there will still need to be years more of tedious work.

Countless painstaking patch clamping experiments of carefully guiding a glass electrode into a single neuron while keeping it alive, just to learn how that one cell type, out of the fly brain's thousands, transforms its inputs into outputs.

Endless sessions of pinning flies under two-photon microscopes, collecting calcium imaging data while the animals walk or groom or navigate an odor plume, slowly building up ground truth measurements of what real brain activity actually looks like during real behavior.

Thousands of hours still to come of building computational models, testing them against that data, failing, and refining them again.

Then, and very likely only then, will there come a day when someone will hit run, and a fly, disoriented in whatever way a fly can be, having been sitting in a vial a moment ago, will find itself somewhere unfamiliar.

It won't know that in the intervening time, it had been anesthetized, embedded in resin, and its brain sliced into thousands of thin sections.

It won't know that those sections were painstakingly imaged, or that its neural architecture was reconstructed from those images, or that thousands of its fellow flies were studied and sacrificed to fill in what images alone couldn't tell us.

It won't know of the billions of dollars and thousands of careers that it took to reach this point or the millions of hours spent staring down microscopes, handling vials, and debugging code.

It will certainly never know that it was once made of proteins and cells and is now made of silicon and mathematics.

It will just beat its wings, lift off, and search for fruit.

This article was narrated by Type 3 Audio for Less Wrong.

It was published on March 19, 2026.

Images are included in the podcast episode description.