The videos from the author are also a lot of fun, and his analysis are quite interesting, highly recommend - https://www.youtube.com/@TheBibitesDigitalLife

Great to see this. Tons more to explore in this space, especially with strides in commodity AI performance.

Best demo I ever saw was ecologist Tom Ray showing Tierra ALife at the MIT AI Lab in 1991. Was a signal for the beginning of a new era, the end of GOFAI and the start of automatic machine trained performance.

Evolution's effectiveness in optimization/efficiency dynamics that Ray first showed in Tierra have been demonstrated in the closed world game players eg AlphaZero but deploying them in open world large scale environments remains open.

That was mentioned at the Linux AI HOwto I think.

In what way does AlphaZero use evolution?

Let it run overnight and the creatures look quite different from where I left it.

Would be interesting if it was easier to identify and compare their lineages. Hard to sort through all these little things swimming around

For the macOS version, I had to manually mkdir The Bibtites.app/Contents/Mods to get it to load.

This is really cool and impressive! It reminds me of something I put together a looooong time ago (https://www.youtube.com/watch?v=6g7pUTWLgEI), although theirs has two magnitudes more effort.

I wonder if they were inspired by it - the style is extraordinarily similar. Skip around the link to see the lil guys.

EDIT:

My memory is shot.

A couple years ago the dev left a comment on that same exact video noting it as an inspiration, and I had already replied to it that it made my day.

The benefits of a bad memory is that my day gets to be made again :)

Just joined as a patreon supporter!

Getting an empty screen with the Linux version.

Same here, Ubuntu 22.10

You need to run it with the —force-vulcan option. The subreddit has details if you search

> stockpiling food in a specific area of the map

Is this emergent behavior or pre-programmed?

Having watched some of the neural network descriptions the game developer has posted, I believe it could easily be emergent if the simulation runs though enough generations.

The neural network can be pre-programmed (and the starting settings offer some basic neural pre-wiring), but has enough sensor inputs and behavior outputs that "If pushing a food pellet, swim to the right at an angle" is a possible behavior. This results in a clockwise swim direction that mechanically nudges other pellets towards the center of the circle.

Other inputs like "swim towards creatures of color #0000FF in your vision cone" (assuming the creature is blue) and "swim towards food in your vision cone" could have the second-order effect of "gathering food towards creatures of the same color -- which are presumably of similar species"

hope author can get money, this absolutely fun.

+1 and inspiring

The Linux version just gives me a dark blue screen with slowly growing memory usage over some minutes.

I've found a workaround on the subreddit, add -force-vulkan as a command line parameter.