So I've been thinking about this crazy idea, and I'm wondering if anyone wants to explore further with me. The idea is to make a videogame that people can play online, involving ants. Instead of simulating what we think ants might sortof be like, as you may remember from the game Sim Ant, I'm brainstorming what we could do with real ants. So imagine ant farms on server racks, with access to some sort of 'game area' outside of the 'ant farm' colony. The game would present live information about the ants and track which ones belong to which colony when they're out in the game area. The game would also have way for the player to influence the ants' behavior, such as food, water, pheromones, predators, etc. Anyway, if you wanna brainstorm on this, lemme know!
Sounds awesome.... but what functions would the human player be able
to control? Obviously you can't directly command an ant to do things.
Perhaps you can dig for them or lay scents?
On Mon, Jul 16, 2012 at 4:02 PM, Valdes <joshval...@gmail.com> wrote:
> So I've been thinking about this crazy idea, and I'm wondering if anyone
> wants to explore further with me. The idea is to make a videogame that
> people can play online, involving ants. Instead of simulating what we think
> ants might sortof be like, as you may remember from the game Sim Ant, I'm
> brainstorming what we could do with real ants. So imagine ant farms on
> server racks, with access to some sort of 'game area' outside of the 'ant
> farm' colony. The game would present live information about the ants and
> track which ones belong to which colony when they're out in the game area.
> The game would also have way for the player to influence the ants' behavior,
> such as food, water, pheromones, predators, etc. Anyway, if you wanna
> brainstorm on this, lemme know!
I love the idea. You can have each player control one ant colony and lay down scents, place food, etc. It could be a battle game where each side has to kill the other (there feels like there's some ethical issues here, no?), or it could be cooperative - getting the ants to collect all of the "whatevers" in the shortest time? Maybe you could actually do it outside... and have capture the flag? Like you have an object which ants want (again, not sure what this is) but has a clearly identifiable signature even underground so you'd know who's colony it was in. Then you try to keep it in your colony for as long as possible, and if it's in someone else's colony you try to induce your ants to go get it...
Probably the first question is to list all the ways you can influence ants, then go from there.
I'm going to come to build night on thursday (finally) with some games to play and stuff, happy to brainstorm -
On Monday, July 16, 2012 4:02:58 PM UTC-4, Valdes wrote:
> So I've been thinking about this crazy idea, and I'm wondering if anyone > wants to explore further with me. The idea is to make a videogame that > people can play online, involving ants. Instead of simulating what we think > ants might sortof be like, as you may remember from the game Sim Ant, I'm > brainstorming what we could do with real ants. So imagine ant farms on > server racks, with access to some sort of 'game area' outside of the 'ant > farm' colony. The game would present live information about the ants and > track which ones belong to which colony when they're out in the game area. > The game would also have way for the player to influence the ants' > behavior, such as food, water, pheromones, predators, etc. Anyway, if you > wanna brainstorm on this, lemme know!
Sounds awesome.... but what functions would the human player be able
to control? Obviously you can't directly command an ant to do things.
I don't know about that. There's a local company, Backyard Brains, that
lets you do just that with roaches. Probably impractical for ant colonies
but still interesting. They've been to AA Makers Faire.
YEA! We could mark the ants of each colony with an identifying dot of
color, or even laser a unique ID QR code on each ant to track their
interactions, feeding, wins, etc. We could make games that explore concepts
of governance and employment. We could induce them to battle, which i see
as more ethical than poisoning ants at the home or office... a bonus of
which is that if we're actually tracking the ants, especially on an
individual basis, it could be useful to science. Individual ants learn and
have been observed to teach each other, including feedback of 'are you
getting it?'. I haven't been to AHA! build night for over a year BUT I do
know it tends to be not a very good environment to focus on anything
thursday nights, due to all the buzz.
-- Valdés
Gamma camera, typically used in a medical setting, tracks isotopes in real time as they travel through the body. You typically either give a patient a sugary food laced with an isotope and watch how the sugar is broken down in cells (usually to detect high metabolism, indicative of cancerous cells), or administer the isotope through an IV.
It would be cool to feed the ants some isotope to track later on. You can get a real time map of where all the ants are and where their food supply is!
Not sure if acquiring a gamma camera or a suitable isotope is feasible, but it would certainly be cool.
> I love the idea. You can have each player control one ant colony and lay down scents, place food, etc. It could be a battle game where each side has to kill the other (there feels like there's some ethical issues here, no?), or it could be cooperative - getting the ants to collect all of the "whatevers" in the shortest time? Maybe you could actually do it outside... and have capture the flag? Like you have an object which ants want (again, not sure what this is) but has a clearly identifiable signature even underground so you'd know who's colony it was in. Then you try to keep it in your colony for as long as possible, and if it's in someone else's colony you try to induce your ants to go get it...
> Probably the first question is to list all the ways you can influence ants, then go from there.
> I'm going to come to build night on thursday (finally) with some games to play and stuff, happy to brainstorm -
> Greg
> On Monday, July 16, 2012 4:02:58 PM UTC-4, Valdes wrote:
> So I've been thinking about this crazy idea, and I'm wondering if anyone wants to explore further with me. The idea is to make a videogame that people can play online, involving ants. Instead of simulating what we think ants might sortof be like, as you may remember from the game Sim Ant, I'm brainstorming what we could do with real ants. So imagine ant farms on server racks, with access to some sort of 'game area' outside of the 'ant farm' colony. The game would present live information about the ants and track which ones belong to which colony when they're out in the game area. The game would also have way for the player to influence the ants' behavior, such as food, water, pheromones, predators, etc. Anyway, if you wanna brainstorm on this, lemme know!
> wrote:
> Gamma camera, typically used in a medical setting, tracks isotopes in real
> time as they travel through the body. You typically either give a patient a
> sugary food laced with an isotope and watch how the sugar is broken down in
> cells (usually to detect high metabolism, indicative of cancerous cells),
> or administer the isotope through an IV.
> It would be cool to feed the ants some isotope to track later on. You can
> get a real time map of where all the ants are and where their food supply
> is!
> Not sure if acquiring a gamma camera or a suitable isotope is feasible,
> but it would certainly be cool.
> On Jul 17, 2012, at 08:26 , Greg Austic wrote:
> I used to play sim ant all the time!
> I love the idea. You can have each player control one ant colony and lay
> down scents, place food, etc. It could be a battle game where each side
> has to kill the other (there feels like there's some ethical issues here,
> no?), or it could be cooperative - getting the ants to collect all of the
> "whatevers" in the shortest time? Maybe you could actually do it
> outside... and have capture the flag? Like you have an object which ants
> want (again, not sure what this is) but has a clearly identifiable
> signature even underground so you'd know who's colony it was in. Then you
> try to keep it in your colony for as long as possible, and if it's in
> someone else's colony you try to induce your ants to go get it...
> Probably the first question is to list all the ways you can influence
> ants, then go from there.
> I'm going to come to build night on thursday (finally) with some games to
> play and stuff, happy to brainstorm -
> Greg
> On Monday, July 16, 2012 4:02:58 PM UTC-4, Valdes wrote:
>> So I've been thinking about this crazy idea, and I'm wondering if anyone
>> wants to explore further with me. The idea is to make a videogame that
>> people can play online, involving ants. Instead of simulating what we think
>> ants might sortof be like, as you may remember from the game Sim Ant, I'm
>> brainstorming what we could do with real ants. So imagine ant farms on
>> server racks, with access to some sort of 'game area' outside of the 'ant
>> farm' colony. The game would present live information about the ants and
>> track which ones belong to which colony when they're out in the game area.
>> The game would also have way for the player to influence the ants'
>> behavior, such as food, water, pheromones, predators, etc. Anyway, if you
>> wanna brainstorm on this, lemme know!
We could try to get ants to interact with sensors that give them godlike environmental controls, such as when the lights turn on and off, temp, humidity, etc. to see what they can figure out, and include new options like that into gameplay. If the ants were individually identified, we could restrict access to food sources to specific colony members to see how that affects their overall organization, as different governmental regimes / economic models that can be instituted.
Maybe at this very moment some superbeing son or daughter is playing with
their "human farm", and doing some of the things you mentioned in this
email:) playing with the barack or romney colonies.
>> YEA! We could mark the ants of each colony with an identifying dot of
color, or even laser a unique ID QR code on each ant to track their
interactions, feeding, wins, etc. We could make games that explore concepts
of governance and employment. We could induce them to battle, which i see
as more ethical than poisoning ants at the home or office... a bonus of
which is that if we're actually tracking the ants, especially on an
individual basis, it could be useful to science. Individual ants learn and
have been observed to teach each other, including feedback of 'are you
getting it?'. I haven't been to AHA! build night for over a year BUT I do
know it tends to be not a very good environment to focus on anything
thursday nights, due to all the buzz.
Maybe at this very moment some superbeing son or daughter is playing with
their "human farm", and doing some of the things you mentioned in this
email:) playing with the barack or romney colonies.
>> YEA! We could mark the ants of each colony with an identifying dot of
color, or even laser a unique ID QR code on each ant to track their
interactions, feeding, wins, etc. We could make games that explore concepts
of governance and employment. We could induce them to battle, which i see
as more ethical than poisoning ants at the home or office... a bonus of
which is that if we're actually tracking the ants, especially on an
individual basis, it could be useful to science. Individual ants learn and
have been observed to teach each other, including feedback of 'are you
getting it?'. I haven't been to AHA! build night for over a year BUT I do
know it tends to be not a very good environment to focus on anything
thursday nights, due to all the buzz.
The hard part i want to meet to brainstorm, is what kind of gameplay story
we can wrap the behaviors in. So the player is an alien, set out to alter
these primitive civilizations it finds in it's travels. Maybe each
civilized planet could be a different ant species.
There could be ant 're-education' rooms, where the computer trains an
individual to behave in some strange way that meshes with different
training given to other ants. Maybe we could use hormonal/nutritional
influences or simple counting and plucking to manage the number of ants in
each caste. Maybe we could give them drugs or strange growth factors. Maybe
we could order genetically modified ants off the internet. I know we can
order viruses with custom genes to insert into our ants for less than the
cost of a laptop.
-- Valdés
It might be easier to start with something simpler.
For example, design a robot-ant ISA with operations similar to the Logo
Turtle.
- Move distance.
- Turn degrees.
- Drop-pheromone data.
- Smell return-pheromone-data.
- When-meeting-another-ant-execute procedure-name.
- Pick-up-food energy.
- Drop-off-food energy.
- Eat-food energy.
And a landscape set.
- Generate landscape. (Places obstacles. Initializes colonies.)
- Rain percent-pheromone-track-lost.
- Drop-food-at location.
- (Some rule set to mediate encounters between rival ants.)
- Step number-of-ant-instructions-to-execute. (Ants with complex or
inefficient logic may be paused and vulnerable.)
The competition would be colony to colony with the mix of ant-roles (nurse,
explorer, gatherer, warrior) representing the each colony's strategic
position. Each move of an ant would consume some amount of energy depending
on the size of the ant. There would be a trade-off between ant size (body
length), speed (in body lengths per step).
On Tue, Jul 17, 2012 at 4:57 PM, Valdes <joshval...@gmail.com> wrote:
> We could try to get ants to interact with sensors that give them godlike
> environmental controls, such as when the lights turn on and off, temp,
> humidity, etc. to see what they can figure out, and include new options
> like that into gameplay. If the ants were individually identified, we
> could restrict access to food sources to specific colony members to see how
> that affects their overall organization, as different governmental regimes
> / economic models that can be instituted.
Soooo maybe the game itself could be entirely computer graphics, but the
actual motions of aliens in these space civilizations could be still based
on tracking real ants. That way the ants wouldn't look all the same to a
casual observer, ie. give each individual a distinctive appearance. Then
significant ant behaviors that aren't obvious to people who don't study
ants can be made more explicit. So for instance, if 2 ants briefly touch
mouths, which they do to distribute half chewed food and water throughout
the colony, it could be made much more dramatic in the game. That way more
players could be on a single ant colony at a time, because there'd seem to
be more going on when each little interaction is embellished.
Maybe networks of interactions between individual ants in the colony can be
graphed into a sort of social graph, where ants that interact with each
other more often will show up visually as clusters. Then the game could
pretend like that cluster is some interest group or company, possibly
related to common colony functions or caste. Then the player could try to
take actions to split the group, or get it to connect to another group
(using pheromone, or food, or a little robot that picks up individuals and
moves them near each other repeatedly, or whatever)
all i know is i should stay away from whatever comes of this. in simant, i
always took a horde of ants and killed anything i could, most notably the
neutral caterpillar and the not so neutral spider. i would not do well
with this new found power.
On Wed, Jul 18, 2012 at 3:34 PM, Joshua Valdés <joshval...@gmail.com> wrote:
> Soooo maybe the game itself could be entirely computer graphics, but the
> actual motions of aliens in these space civilizations could be still based
> on tracking real ants. That way the ants wouldn't look all the same to a
> casual observer, ie. give each individual a distinctive appearance. Then
> significant ant behaviors that aren't obvious to people who don't study
> ants can be made more explicit. So for instance, if 2 ants briefly touch
> mouths, which they do to distribute half chewed food and water throughout
> the colony, it could be made much more dramatic in the game. That way more
> players could be on a single ant colony at a time, because there'd seem to
> be more going on when each little interaction is embellished.
> Maybe networks of interactions between individual ants in the colony can
> be graphed into a sort of social graph, where ants that interact with each
> other more often will show up visually as clusters. Then the game could
> pretend like that cluster is some interest group or company, possibly
> related to common colony functions or caste. Then the player could try to
> take actions to split the group, or get it to connect to another group
> (using pheromone, or food, or a little robot that picks up individuals and
> moves them near each other repeatedly, or whatever)
One of the suggested chemicals is mint oil. It has the wonderful properties
of repelling ants, wasps, and yellow jackets, it smells nice, and it
doesn't give me headaches.
Ants are really tiny, and I want to track them individually over time. Does anyone know of RFID chips small enough to glue to ants? Or a way to print unique identifiers on them, such as QR codes?
Ants are really tiny, and I want to track them individually over time. Does anyone know of RFID chips small enough to glue to ants? Or a way to print unique identifiers on them, such as QR codes?
You might be able to use colored dots. With just a standard camera setup,
you get only a few bits per dot. You can get much more color information by
taking a picture through a prism. The colors in a dot on the back of an ant
will be spread out in the image, giving a spectrum. Alternatively, you can
take a series of pictures with different colored lights, or through
different filters.
An exotic (probably prohibitively expensive) technology is to use
tuned fluorescent dyes and tuned lasers. You can get a couple dozen bits
into a single dot that way. Each dye responds to only one illumination
color, and glows in one of a handful of different colors. Here is entirely
the wrong stuff, because it has extra stuff to make the dye turn on only in
the presence of some particular chemical
http://www.ebioscience.com/knowledge-center/application/flow-cytometr... The concept is right, you shine a 532nm laser at your dot, and use filters
or prisms to detect glows at 561, 578, 615, 667, 695, and/or 785nm. This
gives you 6 bits of data. Repeat with 3 other illumination colors.
Assuming you can find a source for the dyes, you would have to create a
custom mix for each ant. Ant number 1's mix would have just the first dye,
so it glowed at 561nm, ant number 2 would glow at 578, ant 3 would have
both dyes, and thus glow in both colors. Repeat until you get bored of
mixing dyes, catching, and painting ants.
What size ant are you using? How many dots can you put on a single ant, and
have them visibly distinct?
On Thu, Jul 26, 2012 at 12:32 PM, Valdes <joshval...@gmail.com> wrote:
> Ants are really tiny, and I want to track them individually over time.
> Does anyone know of RFID chips small enough to glue to ants? Or a way to
> print unique identifiers on them, such as QR codes?
On Thu, Jul 26, 2012 at 2:15 PM, Paul Haas <hamj...@gmail.com> wrote:
> You might be able to use colored dots. With just a standard camera setup,
> you get only a few bits per dot. You can get much more color information by
> taking a picture through a prism. The colors in a dot on the back of an ant
> will be spread out in the image, giving a spectrum. Alternatively, you can
> take a series of pictures with different colored lights, or through
> different filters.
> An exotic (probably prohibitively expensive) technology is to use
> tuned fluorescent dyes and tuned lasers. You can get a couple dozen bits
> into a single dot that way. Each dye responds to only one illumination
> color, and glows in one of a handful of different colors. Here is entirely
> the wrong stuff, because it has extra stuff to make the dye turn on only in
> the presence of some particular chemical
> http://www.ebioscience.com/knowledge-center/application/flow-cytometr... > The concept is right, you shine a 532nm laser at your dot, and use filters
> or prisms to detect glows at 561, 578, 615, 667, 695, and/or 785nm. This
> gives you 6 bits of data. Repeat with 3 other illumination colors.
> Assuming you can find a source for the dyes, you would have to create a
> custom mix for each ant. Ant number 1's mix would have just the first dye,
> so it glowed at 561nm, ant number 2 would glow at 578, ant 3 would have
> both dyes, and thus glow in both colors. Repeat until you get bored of
> mixing dyes, catching, and painting ants.
> What size ant are you using? How many dots can you put on a single ant,
> and have them visibly distinct?
> On Thu, Jul 26, 2012 at 12:32 PM, Valdes <joshval...@gmail.com> wrote:
>> Ants are really tiny, and I want to track them individually over time.
>> Does anyone know of RFID chips small enough to glue to ants? Or a way to
>> print unique identifiers on them, such as QR codes?
I just found a HUGE queen! Not sure what species she is yet. I hope she
loves her palace. Nice idea using colors! Maybe i could use colored dots,
and then flash colored LEDs individually in rapid sequence timed to
specific camera frames.
-- Valdés
Peoples, we're pretty lassez-faire about moderating this list but try to
remember there's several hundred people receiving these emails and at this
point I think you've crossed over into "time to setup an ant specific
listserv / tumblr / whatever" territory.
I just don't like my inbox filling up with bug reports, it's a visceral
thing.
On Thu, Jul 26, 2012 at 2:33 PM, Joshua Valdés <joshval...@gmail.com> wrote:
> I just found a HUGE queen! Not sure what species she is yet. I hope she
> loves her palace. Nice idea using colors! Maybe i could use colored dots,
> and then flash colored LEDs individually in rapid sequence timed to
> specific camera frames.
> --
> Valdés
