New World Notes

Space Invaders Meets Darwin: Mini-Game Where Waves of Aliens Pass Genes to Next Generation -- And Evolve

Invaderz is a cool little microgame (space bar to shoot, A/left arrow and D/right arrow to move) which demonstrates some big potential for artificial life in games: It's basically like the classic game Space Invaders, but any time you kill an invading alien, a genetic algorithm passes its genes on to the next generation of invaders. Sometimes those genes mutate, and sometimes those genes make the invaders more dangerous. In other words: The aliens artificially evolve to become better predators.

"Usually in a more traditional genetic algorithm," developer Victor Ribeiro tells me, "A mutation has only a small chance to happen, like 1%." To speed this evolutionary simulation up, he coded the mutations to happen 10% of the time. "Playing the game you might see some enemies that are very similar to another, with just one small difference: the mutation. That small difference will affect how the enemy moves; maybe it will affect it in a positive way, maybe not. The thing is, that mutation is unique."

So over time, they could theoretically evolve to the point where the aliens are better than the human player. (Darwinian theory in action, only with more pew-pew-pew sounds!)

In tests, Victor has seen the aliens evolve to become faster, so "by the time [the bullet] arrives at the enemy's location, the enemy will be long gone."

The code is available on his GitHub if you want to play with it. Seems like this could be a great system to implement into full-fledged simulation games. The thing to keep in mind, as Victor cautions, is that evolution is a long process and most mutations are not a benefit or disadvantage, especially in a game context: 

Watch an AI-Controlled Avatar "Learn" How to Dribble In Real Time - Huge Implications for Games, AR/VR, and Real Life Robotics

What you're looking at above appears to be a pretty big breakthrough in artificial intelligence -- an avatar "learning" how to dribble in real time through AI. DeepDribble is a project from researchers at DeepMind and Carnegie Mellon, and the idea is to demonstrate how an AI-controlled avatar can acquire the complex and subtle set of integrated physical skills and coordination that humans take for granted:

Basketball at its peak is known for intricate ball handling, quick pivoting, and fake outs. A good player will fluidly transition between have a variety of dribbling skills. In order to recreate this, the trained characters need to be able to transition between separately trained behaviors... Libin Liu and Jessica Hodgins use the same methods for training single skills to create a multi-skill control graph (which allows transition between motion fragments of two different skills). The training is done incrementally, behavior by behavior, to ensure integration occurs with proper transitions for each skill. The result is a player that can perform multiple ball handling skills in a variety of orders, changing course with seamless blending.

"First we build a bio-mechanical model of the digital character, using physics to simulate bones and virtual muscles," Deep Motion's Libin Liu tells me. "We put the digital character in the physically simulated virtual world and let it mimic how real players perform the motions using their muscle forces."

The training process, as you might have imagine, requires quite a lot of power -- but Liu says the learning arc took about as long as a single televised NBA game -- or even half of one:

Because of Bento, Rideable Animals in Second Life Now Officially Amazing

I haven't blogged much about rideable Second Life animals since the release of the articulated avatar skeletons of (code-named Bento), but in recent months, they've gotten pretty impressive in their detail and realism. Pictured here is a variation of Tyrian Slade's Water Horse line on the Marketplace here. Comes with countless customization options and many gestures, including Bow, Turn,  Canter, even AFK, and will cost you L$3,889 (i.e. under $16). Which is kind of pricey, but this sucker can move about as gracefully as a mounted horse from, say, Skyrim -- look:

Cat Life Launches: Feed, Brush, Pet, Play With a Virtual Cat in Second Life -- Or Be the Cat Yourself!

Cat Life, the ambitious pet game for Second Life you read about last December, is now finally available in the SL Marketplace right here. Created by the popular virtual animal/NPC brand Zooby (Facebook page here ), lead designer Carrie Tatsu narrates the walk-through tutorial above.

As the name suggests, it's both a pet care game and a game where players can actually be the cat themselves: "I can be a cat and you can be a cat and we can share our accessories and gain tokens which can be redeemed for toys," she says. 

And yes, that means doing pretty much anything an actual cat does, as part of the game:

Guess Which Worm is Real & Which is Virtual in This Turing Test That Made the Royal Society Audience Gasp Out Loud

Since folks have asked, this is the Turing test I showed at the @royalsociety on Tuesday. One of the two is the latest @OpenWorm simulation (c302+Sibernetic) replayed using @GeppettoEngine, the other is a real C. elegans. Can you tell which is which? pic.twitter.com/g65aR4E4Bl

— Matteo Cantarelli (@tarelli) February 1, 2018

Here's a cool follow-up video to last week's post on OpenWorm's presentation to the prestigious Royal Society: "This how Matteo Cantarelli's talk at the Royal Society ended with a visible gasp in the audience," Giovanni Idili‏ tells me. I like that Matteo calls it a "Turing Test", which I guess is the right way to describe a virtual simulation of a code-based worm. (Also because OpenWorm is publishing a paper in the Royal Society’s Philosophical Transactions B, whose past contributors literally include Alan Turing.)

Anyway, make your guess, then find the correct answer after the break.

Open Source Virtual Life Project Successfully Presented at the Legendary Royal Society Scientific Body

Remember the OpenWorm Project that I wrote about a few years ago? It's a collaborative, open source attempt to construct an artificial life form from the cellular level to the point where it's able to have basic problem-solving abilities. This week the project (like the worm, based on the c. elegans nematode) evolved to a new milestone, when team leaders got to present their work at the Royal Society in London. Yes, the same Society whose fellows include Isaac Newton and Stephen Hawking. The meeting was organized by OpenWorm’s scientific advisor John White (father of the connectome neural map), a fellow of the Royal Society, and according to Open Worm co-founder Giovanni Idili, it was a rousing success:

“It was a two days back to back conference with loads of c. elegans modelers and experimentalists with OpenWorm talks peppered in, a meeting first of its kind with some great work presented and generating a lot of discourse between communities that historically don’t talk to each other much (biologists and computational modelers)," he tells me. "The Royal Society is gonna publish a special issue with a paper for each of the talks too.”

Specifically, this paper on virtual life will be published in a future issue of the Royal Society’s Philosophical Transactions B, whose past contributors include Benjamin Franklin, Michael Faraday, and Alan Turing. So yes, this is pretty impressive.

“Kind of a big deal for us to go from scrappy open source, open science project to hosting a meeting at the Royal Society,” as Idili puts it. “But we’re still scrappy and proudly so!” (Speaking of open source, one of Open Worm's first contributors is John Hurliman, an OpenSim pioneer.) 

While I wrote “virtual life” up there, by the way, the creators are a bit more cautious around that term:

Coming Soon to Second Life: Bento Mesh Cat With Emotive Expressions, Pet-Raising Game, and Both Avatar & NPC Functionality

Here's a pretty impressive preview of Cat Life, a new pet game coming soon to Second Life, perhaps as early as next month. Created by the well-known virtual animal/NPC brand Zooby, you can follow updates by liking the company's Facebook page.

Zooby's Carrie Tatsu narrates this preview, and as you can tell from the video, this cat uses the new Bento technology for highly expressive avatar skeletons.  Carrie tells me the pet-raising game will be just as complex:

"The caring system and token system can be used to receive cat toys," she tells me. "And [there's] potential to ‘gamify’ more with add ons such as cat tag, or cat yoga or cat fishing." Perhaps even cooler than that, "you can be a cat and I can be a cat and we can share our accessories and buy each other litter and food and you can eat from my bowl and I can use your cat nip. Or litter box." 

Yes: Be a virtual cat and poop together with friends.

And once the Animesh system for creating NPCs is released to the official Second Life grid, get ready for a full-fledged mini-MMO for virtual cats:

Ask Developer of Popular Meeroos & MadPea Brands About Building Games & Businesses in Second Life

Really interesting Ask Me Anything thread with Levio Serenity, a developer of the hugely successful Meeroos pet game and for MadPea's ambitious adventure games in Second Life. Turns out Levio has been a programmer for Fortune 500 companies who jumped at a chance to realize a childhood dream of designing games, starting with Meeroos, one of the first huge virtual pet game hits in Second Life (despite notorious griefing incidents).  It's got great insights into running successful businesses in Second Life, because he's answering in-depth questions on the technical details -- such as this one on using Linden Script Language versus an external server to run the actual game code:

Creators of Open Source Virtual Life Project Presenting to the Royal Society

Remember the OpenWorm Project that I wrote about a few years ago, a collaborative, open source attempt to construct an artificial life form from the cellular level to the point where it's able to have basic problem-solving abilities? The project (like the worm) has evolved since then, to the point where the creators have been invited  to present their work to the Royal Society:

This meeting will bring together neuroscientists, physicists, and engineers to discuss advances in neural activity imaging, behaviour quantification, and multiscale simulations and how they are bringing the goal of whole-animal modelling at cellular resolution within reach.

If the group sounds vaguely familiar, it could be because its past and present fellows include obscure scientists such as Isaac Newton and Stephen Hawking.

"We are super proud of this because starting as an open science/open source project as a couple of guys in a hangout being recognized by such an established organization is pretty massive for us," Open Worm co-founder Giovanni Idili tells me. "Also they'll publish an issue based exclusively on this meeting we are organizing so that's pretty exciting and dare I say unprecedented." (Maybe there's another legendary scientific body that's devoted attention to the first digital life form of its kind,  but I'm guessing not.) 

The Royal Society presentation happens next January. Meantime, says Giovanni, the worm keeps evolving:

Play Clusters: Artificial Life Simulator for iOS Modeled on Gaia Hypothesis Theories, Created by MMO Pioneer Jeffrey Ventrella

Clusters is an ambitious new artificial life game/toy in the iOS App Store created by Jeffrey Ventrella, a developer who did some pioneering work for Second Life and other virtual worlds. As you might imagine, his version of an artificial life simulator is equally ambitious.

“In the spirit of classic artificial life models, such as the Game of Life,” says Jeffrey, “Clusters doesn't literally mimic life forms, but it does echo certain fundamental dynamics which are the basis of all life, such as self-assembly, sensitivity to environmental conditions, self-similarity, and dynamic equilibria.”

Clusters is based on the theories of biologist Lynn Margulis, which run counter to Darwin’s competition-based model: “She showed that cooperation and symbiosis are just as important, and perhaps even more important. She championed the theory of endosymbiosis, which claims that early organisms were so mutually dependent that they actually fused their identities, and became more complex organisms. And this is what made complex life possible on Earth.” Margulis later incorporated these ideas into the famed Gaia Hypothesis she helped James Lovelock develop, ”which claims that the Earth is an organism, with its own self-regulating dynamics… it gives power to the idea that we are all connected.”

Fricking heady stuff for an iOS app! Unsurprisingly, emergent patterns occur when Clusters runs: