http://vimeo.com/36579366
This is an interesting perspective to present. It's also a great way to showcase some very inventive concepts.
Much to think about.
Wednesday, October 31, 2012
Monday, October 29, 2012
Yet more GA ideas
http://burakkanber.com/blog/machine-learning-genetic-algorithms-part-1-javascript/
Interesting article above. Stired an idead that has been bubbling around for ages.
The issue I see is that the chromosome of the GA and the fitness function are seperated. The choromosome is mutated, mated and selected, while the fitness function is static. This treats the "chromosome" simply as "data" while the fitness function, mutation function, mating function and selection function act as immutable "code".
I feel like I have been in this conversation before, but I want some way to load all the functions into the chromosome and allow both the "code" and the "data" to mutate and be mixed up at the same time.
For this to happen, there has to be a simple generic machine to parse the data in the chromosome and interpret some of it as "payload", some as fitness function, some as mating function, some as selection function and some as mutation function.
Chances are that this model will cause catastrophic failure in most scenarios... but its a fun idea.
By allowing all these functions to mutate and become part of the GA, it should make for a more flexible system, that does not encode any of the assumptions of the designers into it. It should allow the system to evolve without any unknown constraints.
What will this need?
A set of abstract mechanisms to iplement the various functions (Fitness, mutation, selection, mating) at a population level along with a chromosome of "data".
The simplest place to start is halfway.... Start with a library of fitness functions and use a simple flag in the data chromosome to pick which one to use on the GA. (Theoretically, this will pick the fitness function which gives the highest "score" for different families of GA's.
Ditto for mutation, selection and mating functions... just sit around and come up with every weird and amusing function possible, then let evoluation selecte the winners.
Interesting article above. Stired an idead that has been bubbling around for ages.
The issue I see is that the chromosome of the GA and the fitness function are seperated. The choromosome is mutated, mated and selected, while the fitness function is static. This treats the "chromosome" simply as "data" while the fitness function, mutation function, mating function and selection function act as immutable "code".
I feel like I have been in this conversation before, but I want some way to load all the functions into the chromosome and allow both the "code" and the "data" to mutate and be mixed up at the same time.
For this to happen, there has to be a simple generic machine to parse the data in the chromosome and interpret some of it as "payload", some as fitness function, some as mating function, some as selection function and some as mutation function.
Chances are that this model will cause catastrophic failure in most scenarios... but its a fun idea.
By allowing all these functions to mutate and become part of the GA, it should make for a more flexible system, that does not encode any of the assumptions of the designers into it. It should allow the system to evolve without any unknown constraints.
What will this need?
A set of abstract mechanisms to iplement the various functions (Fitness, mutation, selection, mating) at a population level along with a chromosome of "data".
The simplest place to start is halfway.... Start with a library of fitness functions and use a simple flag in the data chromosome to pick which one to use on the GA. (Theoretically, this will pick the fitness function which gives the highest "score" for different families of GA's.
Ditto for mutation, selection and mating functions... just sit around and come up with every weird and amusing function possible, then let evoluation selecte the winners.
Monday, October 22, 2012
Beautiful Review of Dishonored
http://www.gamasutra.com/view/news/178604/Why_people_are_playing_Dishonored.php
This review is verging on over-literary but manages to avoid being sickening. Its just a beautifully lyrical celebration of something that the reviewer seems to enjoy deeply.
Very nice writing.
This review is verging on over-literary but manages to avoid being sickening. Its just a beautifully lyrical celebration of something that the reviewer seems to enjoy deeply.
Very nice writing.
Paper on Legal Rights for Robots
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2044797
Citation
Darling, Kate, Extending Legal Rights to Social Robots (April 23, 2012). We Robot Conference, University of Miami, April 2012. Available at SSRN: http://ssrn.com/abstract=2044797 or http://dx.doi.org/10.2139/ssrn.2044797
http://io9.com/kate-darling/
http://www.escapistmagazine.com/news/view/119440-MIT-Researcher-Proposes-Rights-for-Robots
http://lsolum.typepad.com/legaltheory/2012/03/conference-announcement-we-robot-at-miami.html
This is a really interesting paper that explores the rationale and implications of granting rights to social robots.
It opens a whole can of scenarios that legislators will need to wrestle with. Some of the obvious edge cases that spring to mind are:
* How might these rights apply to the manufacture, repair and disposal of the robots?
- Manufacturer's need to "test" robots, which may involve testing them to destruction, or testing them in non-legal ways to ensure saftey for the user.
- Repair technicians need to disassemble and replace components and may place the robot in a state that is "illegal". They may also return a robot to the owner in a "functional" state but a state of "change" which the owner finds disturbing.
- How do you dispose of a robot if it has rights? Can you put it out for the trash? Can you bury it? How do you dispose of a robot that is damaged or completely non-functional?
* One scenario that was mentioned in the paper as being a potential area for legislation is the domain of sexual behaviour with robots. In the case where the behaviour crosses the boundaries of currently socially acceptable behaviour, should this also apply to robots?
The complication here is that a robot can be created specicially for the purpose of either the practice or theraputic behaviour modification therapy around these practices. As the robot becomes a "tool" in this scenario and may be programmed in ways that the practice does not engender "pain" but instead activly "enjoy" the behaviour. Or engage in a way that theraputically modifies the behaviour in the human participant.
In this kind of scenario, the legal protection of the robot may complicate therapies that are designed to benefit society.
* Scenarios where robots could or are Injured or destroyed through use are complicated. It's quite reasonable to conceive of robots being inadvertantly injured during "work" such as dropped or broken through rough play. In these cases, would there be a requirement for "repair" or trauma councelling for the affected humans?
* What about indirect attacks on humans by people attacking their social robots? Similar to cases of companion animals being attacked by other animals or humans to inflict emotional harm to the owner.
Things to think about.
The obvious extension of this is about purely software artifacts. Social Software Agents. Would they also be granted similar rights? Especially where they embody some unique content or learning that has gone on to shape their behaviour with that of their human user/owner and have a strong bond. Do they have some right of protection, existance, power?
The future will be interesting.
Friday, October 19, 2012
Exhaustion....
I have decided that the thing that will kill the movie industry is simple inertia. Changing market dynamics, movie pirates, old school business models, more TV channels, too much Internet, gaming etc... all these are just symptoms and distractions. The root cause is the incredibly slow pace of production.
The "industry" (and remember its a collection of individuals with no common purpose; so herd or rabble may be a better description) generates a finite number of movies per annum. Of these, only a certain number will be of interest to any particular group of movie viewers.
Viewers come in two groups, the first is "Indiscriminate Viewers" who will watch anything as the opportunity arises. The second is "discriminating viewers" who are in some fashion selective in their viewing habits. These are people who will refuse to watch some movies given the opportunity and further will expend effort to seek out movies that satisfy their interest.
Generally, I would suggest that over time an indiscriminate viewer will move toward being a discriminate viewer; simply though the mechanims of experience. They find what they enjoy and they seek to maximise that.
So, given that the longer a viewer watches movies, the less (overall) of the annual production they will likely consume.
For instance, a viewer who finds they are a fan of romantic commedy, may seek to see a movie a week (its a figure, don't worry too much about it) which means they want to see 52 romantic comedies a year. What do they do if they can't find something to watch? (Go elsewhere, do something else... probably) The point is that they cannot point this out to the industry... cause the industry is a figment of the collective rabble. Keep in mind that romantic comedy is a very well represented genre in the movie back catalog.
In the event that we look at a fan of a smaller genre...say sci-fi thrillers. Then they may be offered a new release bi-yearly... if that. How does "the industry" (fictional group... I know) expect to keep that persons attention? They can keep up with the new releases if they have a move marathon once every decade... yah!
The other thing is that they can work through the existing back catalog of all sci-fi thrillers ever made in a couple of weeks of casual viewing. Job done. Genre exhausted.
"The Industry" has simply failed them. They may go and find something else to entertain themselves... but its random chance if that thing happens to be another genre of movies. Its just as likely to be any other entertainment... sport, games, porn, fine art restoration, busking.... racing goats... whatever. The odds are against the movie industry getting them back as a paying customer.
My point? Ok... I think the thing that has spoilt the party is simply access to databases of movies. Its now possible to search through the genre that you are interested in and tick off all the ones you have seen, all the ones you would like to see and any that you never want to see...and get a survey of the scope of the situation. There is no longer any mystery. No hope that there is one out there that will keep you interested and searching. You can be sure.
I've done this with the Genres that I'm interested in. I can see the end of my movie watching fast approching. I have certainty that I'm done! Now its just maintenance mode. As new movies come out, I can catch up with them in my own time and tick them off. In between I can re-watch some of the good ones... but essentially there is not a lot left for me. I have thoughly checked all the other genres and frankly, they're just not interesting. Chances are that I am un-representative of the rest of the movie watching population... but as it gets easier and software takes over helping people to organise their media collections... I expect this "ticking off" exercise to be simpler. The only hope is that there will be some feedback to the industry to give them (collectivly) a better idea of the genres and the frequency of delivery that will hold their audience.
After a lifetime of regular cinema and movie watching.... its a bit sad to see the end of it. But unless things change, there's really only one movie yet to be released that I am looking forward to and thats it.
TV Series are a different matter. I think they are filling the space between movies for a lot of the genres. The problem with TV Series is simply the commitment from a production company is so much greater. This makes it harder to get niche genre series off the ground. I wonder what the failure dynamics are with niche series? I wonder if becuase its in a less crowded niche that there is less risk of failure? Interesting.
I am a bit vauge as to where I was going with this whole post... probably just that the movie industry is too slow to release anything and that access to databases has removed the uncertainty about the movies that are availible.
The "industry" (and remember its a collection of individuals with no common purpose; so herd or rabble may be a better description) generates a finite number of movies per annum. Of these, only a certain number will be of interest to any particular group of movie viewers.
Viewers come in two groups, the first is "Indiscriminate Viewers" who will watch anything as the opportunity arises. The second is "discriminating viewers" who are in some fashion selective in their viewing habits. These are people who will refuse to watch some movies given the opportunity and further will expend effort to seek out movies that satisfy their interest.
Generally, I would suggest that over time an indiscriminate viewer will move toward being a discriminate viewer; simply though the mechanims of experience. They find what they enjoy and they seek to maximise that.
So, given that the longer a viewer watches movies, the less (overall) of the annual production they will likely consume.
For instance, a viewer who finds they are a fan of romantic commedy, may seek to see a movie a week (its a figure, don't worry too much about it) which means they want to see 52 romantic comedies a year. What do they do if they can't find something to watch? (Go elsewhere, do something else... probably) The point is that they cannot point this out to the industry... cause the industry is a figment of the collective rabble. Keep in mind that romantic comedy is a very well represented genre in the movie back catalog.
In the event that we look at a fan of a smaller genre...say sci-fi thrillers. Then they may be offered a new release bi-yearly... if that. How does "the industry" (fictional group... I know) expect to keep that persons attention? They can keep up with the new releases if they have a move marathon once every decade... yah!
The other thing is that they can work through the existing back catalog of all sci-fi thrillers ever made in a couple of weeks of casual viewing. Job done. Genre exhausted.
"The Industry" has simply failed them. They may go and find something else to entertain themselves... but its random chance if that thing happens to be another genre of movies. Its just as likely to be any other entertainment... sport, games, porn, fine art restoration, busking.... racing goats... whatever. The odds are against the movie industry getting them back as a paying customer.
My point? Ok... I think the thing that has spoilt the party is simply access to databases of movies. Its now possible to search through the genre that you are interested in and tick off all the ones you have seen, all the ones you would like to see and any that you never want to see...and get a survey of the scope of the situation. There is no longer any mystery. No hope that there is one out there that will keep you interested and searching. You can be sure.
I've done this with the Genres that I'm interested in. I can see the end of my movie watching fast approching. I have certainty that I'm done! Now its just maintenance mode. As new movies come out, I can catch up with them in my own time and tick them off. In between I can re-watch some of the good ones... but essentially there is not a lot left for me. I have thoughly checked all the other genres and frankly, they're just not interesting. Chances are that I am un-representative of the rest of the movie watching population... but as it gets easier and software takes over helping people to organise their media collections... I expect this "ticking off" exercise to be simpler. The only hope is that there will be some feedback to the industry to give them (collectivly) a better idea of the genres and the frequency of delivery that will hold their audience.
After a lifetime of regular cinema and movie watching.... its a bit sad to see the end of it. But unless things change, there's really only one movie yet to be released that I am looking forward to and thats it.
TV Series are a different matter. I think they are filling the space between movies for a lot of the genres. The problem with TV Series is simply the commitment from a production company is so much greater. This makes it harder to get niche genre series off the ground. I wonder what the failure dynamics are with niche series? I wonder if becuase its in a less crowded niche that there is less risk of failure? Interesting.
I am a bit vauge as to where I was going with this whole post... probably just that the movie industry is too slow to release anything and that access to databases has removed the uncertainty about the movies that are availible.
Sketching out an intelligent Audio File Manager...
Itunes has implemented a metadata database with references to the files. This removes the need to use the fileing system as a simple database and store meta data in tags within the file itself.
The question is what is immutable and what is mutable meta-data.
Name of the track, length, encoding details... artist, which album it was released on, year of release... that kind of thing. These are immutable (to a certain degree)
The mutable meta data is more about how the user wants to arange and manipulate their music collection. Groups, genres, tags, play lists, mood ratings, star ratings... etc. How you want to sort, search and filter the items in the library.
Since this set of structures is not unique, clean or complete... its cannot be implemented over the file system+tag structure in use by the majority of music managers. I also think that iTunes is fundamentally flawed because it enforces the distribution and packaging model favored by music retailers. The Album. This is hard wired into its DNA in a way that forces consumers to still consider it as a meaningful entity in the structure. It's neither valid for the music artist nor for the consumer. It's an artifact of last century.
It's now as valid to subscribe to an artists channel. Treat them as an entertainment source that the consumer tunes into. Allow the artist to produce their ideal playlist. Allow friends to trade play lists like we used to trade mix tapes. Its what I want to listen to ... how I want to hear it.
Mashups are the next level of sophistication for the music listener. Again similar to a mixtape... but with the possibility of interactive scripting. Put branch points and structure in a play list. Let it adapt the volume and tempo to suit the mood of the listener. Call it a MixList.
Develop a killer mixlist for listening to over dinner... then once the kids are in bed... change the groove for some quiet time... or rev your engine in thats how you roll. Having your own personal channel that feeds from a range of sources personalises it to whats going on, how you are feeling and where you want the music scape to take you...
Now how to make the beast? Seems simple enough... but is this a media manager? Its more like an AI driven media player.
For a media curator... well thats something different. This is simply a tool. It cleans up and maintains your library. All the basics like filing and taking out the garbage. It could also troll for new material. Automatically download based on preferences. Slide them into your music time when you are in the mood to experiment and see how you respond.
The next level is developing an engine that can analyse the music itself and do some sorting and categorising for you... and then build this information into the sound feed.
The question is what is immutable and what is mutable meta-data.
Name of the track, length, encoding details... artist, which album it was released on, year of release... that kind of thing. These are immutable (to a certain degree)
The mutable meta data is more about how the user wants to arange and manipulate their music collection. Groups, genres, tags, play lists, mood ratings, star ratings... etc. How you want to sort, search and filter the items in the library.
Since this set of structures is not unique, clean or complete... its cannot be implemented over the file system+tag structure in use by the majority of music managers. I also think that iTunes is fundamentally flawed because it enforces the distribution and packaging model favored by music retailers. The Album. This is hard wired into its DNA in a way that forces consumers to still consider it as a meaningful entity in the structure. It's neither valid for the music artist nor for the consumer. It's an artifact of last century.
It's now as valid to subscribe to an artists channel. Treat them as an entertainment source that the consumer tunes into. Allow the artist to produce their ideal playlist. Allow friends to trade play lists like we used to trade mix tapes. Its what I want to listen to ... how I want to hear it.
Mashups are the next level of sophistication for the music listener. Again similar to a mixtape... but with the possibility of interactive scripting. Put branch points and structure in a play list. Let it adapt the volume and tempo to suit the mood of the listener. Call it a MixList.
Develop a killer mixlist for listening to over dinner... then once the kids are in bed... change the groove for some quiet time... or rev your engine in thats how you roll. Having your own personal channel that feeds from a range of sources personalises it to whats going on, how you are feeling and where you want the music scape to take you...
Now how to make the beast? Seems simple enough... but is this a media manager? Its more like an AI driven media player.
For a media curator... well thats something different. This is simply a tool. It cleans up and maintains your library. All the basics like filing and taking out the garbage. It could also troll for new material. Automatically download based on preferences. Slide them into your music time when you are in the mood to experiment and see how you respond.
The next level is developing an engine that can analyse the music itself and do some sorting and categorising for you... and then build this information into the sound feed.
Subscribe to:
Posts (Atom)