The fly window

REPLAY · recorded 2026-09-16

This is a computer model, not a living fly.

The clock is the model's clock, not a real fly's.

Can an added picture link change another model?

I used the response of a sending model to make the next picture for a receiving model.

What the fly's eye sees · How the model's cells respond

A still from the saved run while the movie gets ready

Left, what the fly's eye sees. The picture is drawn on the eye's 721 facets, which is why it looks grainy. Where the picture comes from: a picture we made.

Right, how the model's cells respond. Each block is one group of cells of the same type. Brighter means responding harder. The blocks are a list, not a brain shape. What the blocks show: the model's answer.

REPLAY means you are watching a saved run. Nothing is running live now. The date is when the run was saved.

Top: sending model. Bottom: receiving model. In each row the input is left and the response is right. These are computer models, not living flies.

The picture is paused.Pause stops the saved run. Play starts it again from the same moment.

A mixed-up link changed the receiving model almost as much.

This cannot tell us what a living fly remembers or feels because it uses a computer model. The remaining ordinary explanations are listed with the measurements.

Next: Compare links with mixed-up times and matched picture strength.

EXPLORATORY · An outside reviewer checked the numbers on this card and asked for plainer wording. The result is still exploratory.

Fly One is a computer model of the parts a fly uses to see. It is not a living fly.

Link to this saved record

See the measurements

Watch another test

Some tests have several views. A different view is not a new experiment.

What is still open?

These are questions about the tests. They are not steps toward proving that a model has feelings.

Uncertain

After a picture ends

The early burst is saved, but no lasting trace passed the later check. An ordinary brief wobble is still a possible cause.

Read the saved test
Uncertain

Signals sent back

Weaker links back to earlier cells change the model. The test still needs to separate link direction from link strength.

Read the saved test
Not tested

Working together across a brain

These tests did not measure a loss of shared response across a whole brain.

Not tested

The main navigation hub

Turning off a sight-cell group changed the model. It did not test the main navigation hub of a fly brain.

Read the saved test
Not tested

Sleep and drugs

No sleep or drug experiment was run. Changing a model link cannot establish either state.

Tested

Added picture links

The link delivered a picture and changed the receiving model, but no setting passed both required comparisons.

Read the saved test
Uncertain

Picture speed

A slow change at the start of a short clip affected the speed reading. Attention was not tested.

Read the saved test