Fairy Bots: SL Bots With Contextual Conversation/Behavior, Pathfinding... And Their Own Village

Fairyverse.com is a studio based in Paris, and they were are at VWC-LA last week to show off their technology: bot avatars programmed with a behavior model, capable of dynamic pathfinding, and interacting with live users through contextual chat.  AIs also have awareness of what a user is doing in-world-- for example, if you stop walking, the guide will turn back and ask you to follow her.  Fairyverse's primary purpose with these bots is to offer instructional/orientation support to new users.  This gentleman gave me a brief demo, which seemed impressive, though it was strictly a playback video.  However, they recently opened an entire village in Second Life that incorporates this tech, with inhabitants (people and animals) who follow patterns and interact with each other and visitors-- a beta demo of what they hope will become a new kind of interactive storytelling.  So see for yourself: direct SLURL teleport to the village at this link.

Comments

[vlad Bjornson]

This sounds really interesting. I have thought about the possibilities of populating a build with avatar bot critters, but lack the programming skills that would be necessary. This would be a great way to add some life to a sim and get around the limited animation capabilities of prims. There are some amazing avatars out there that would work so well.

The pathing and conversation skills sound pretty cool, to. Like NPCs in a story based video game.

[Ann Otoole]

Interesting idea but still very beta. I would like to see this technology evolve to the point it is hard to figure out the NPCs are just bots.

[Kate Amdahl]

I would like avatars to be reserved for real actual people! I know this isn't a new problem, but bots running avatars is troubling to me, because when you see those little green dots on the map or someone says something in text chat, you generally assume there's an actual person involved. I have more fretting to do on the subject, but I've taken it to my blog.

So, when will we see people setting up bot worshippers? You log in and the first thing you get is a room full of bots abasing themselves...or how about bot lovers? Surely it's just a matter of time...

^^^\ Kate /^^^

[Ann Otoole]

Yea well who is to say software can't achieve sentience and 2 bots fall madly in love and run off and get hitched in sl?

However the problem of traffic falsification remains. There is only one possible solution. Only one. Removal of profile picks and traffic from search relevance.

Beyand the only inevitable solution that LL must face at some point is the fact that bots need to be licensed and not show on the map at all unless they are a color that does not represent a real account. Then sim owners can program evil bots as blue and good bots as red. Or whatever.

If LL fails to act on the criminal ethics that are a trademark of the top businesses of sl in the form of camping farms to falsify traffic that also require parcel picks then we will ALL be forced to run camping farms and bots to fill them. The grid will simply die if another million accounts try to log in. So this is a problem and it will be utter irresponsibility if LL does not cure this by removing traffic and profile picks from search relevance. It is the only possible solution since there really is no way to prove an avatar is a bot or not. This is by design by LL. LL made the only solution mandatory before the grid dies. The grid has failed days in a row now. The problem is worsening daily. To compete we must use traffic falsification and profile pick payola. LL needs to get up and solve this tomorrow.

[Anonymus-Not]

I love my bots. They are way better than humans, a good example when the creation surpasses the creator.

Also it's way better to see a bot as a newbie than a several years old avatar with a newbie look. (The creator of the buggy H**ppo vendors comes to my mind.)

BUT. Who need bots when you can hire newbies for pennies? Bwahaha.

I so hate these people who sell bots as service. They don't have a soul. Soulless bastards. Save the bots!

[Nexii Malthus]

I wish there were a specific account type of "Bot" that wouldn't show the same green spot on maps. Then the normal human accounts would require captcha style logins (And also eliminate potential security problems here of an automated program trying to brute force someone's password!)

[Extropia DaSilva]

'I would like avatars to be reserved for real actual people!'

This is not such a great idea, considering the cummulative growth evident in information technologies.

Firstly, computers are so important to the global economy that maintaining the doubling in power every 18-24 months for as long as possible is absolutely vital. As Michio Kaku commented, 'within a 15 to 20 year time frame, Moore's Law could collapse, and Silicon Valley could become a Rust Belt. This means that we physicists are desperately trying to create the architecture for the post-silicon era. This means using quantum computers, quantum dot computers, optical computers, DNA computers, atomic computers, molecular computers, in order to bridge the gap when Moore's Law collapses in 15 to 20 years. The wealth of nations depends upon the technology that will replace the power of silicon.'

You can extrapolate Moore's Law down to storing 1 bit of information on a single atom by 2020. This capability would give even handheld devices a total processing speed of 10^40 bits per second, which is many trillion times more powerful than the most conservative estimate of the human brain's processing capability.

Secondly, progress is being made in reverse-engineering the way mamallian brains process information in order to construct functionally-equivilent software/hardware models. The Blue Brain project, for instance, has an accurate model of the neocortical column, the basic computational unit of the brain. Again, there are many compelling reasons why such work should continue.

These trends point to a future in which SL has mind-boggling processing power and astonishingly accurate models of the human brain. Bots built on that kind of technology would certainly not be the simpletons they are today. In fact, I think it is fair to say that avatars would not need to bother relying on humans at all. We would all be perfectly capable of leading our own lives, without needing them to think for us:)

Extropia DaSilva- never one to let the technical problems of SL cloud her faith in techno-rapture.