Archive for November, 2008

False Beliefs

I’ve been reading cognitive psychology this morning trying to get my projects done for the end of the semester. I ran across an interesting experiment along the lines of Piaget’s conservation of volume.

In the false belief experiment a child is shown a box full of candy while two experimenters are in the room. One of the experimenters leaves and the other replaces the candy in the box with pencils and closes the box. When the experimenter returns the child is asked what the absent experimenter thinks is in the box.

Up until the age of three, children will think that the absent person believes the box to contain pencils because they can’t yet understand that what they know is not somehow also known by the other person as well.

There’s some interesting work at the Rensselaer Artificial Intelligence and Reasoning Laboratory on an architecture that reproduces the development of false belief in artificial intelligences in Second Life.

In the first example, the AI fails to understand that the absent party has a false belief:

And in the next, an AI modeling an adult does have that ability:

Leave a Comment

Iterative Science Experiments

In my computational modeling class on Friday, Dr. Palmeri discussed a type of iterative experimentation put forth by Jay Myung & Mark Pitt at Ohio State.

The basic work of the course is models of the mind that make predictions of human behavior. Specifically, at this point we are looking at when there are competing models and researchers wish to distinguish which is the best able to produce realistic data.

The example being used was the discrimination of memory retention models and the time points at which to test for retention so as to optimally distinguish between models. The process starts off and as data is collected, best-fitting values for the parameters of the models can be calculated.

Given those parameters then, there are memory retention test points where the competing models will be expected to give the most divergent predictions. The data collection algorithm then can adjust the experimental setup to collect at those points.

The data from those points, is then added to the dataset that distinguishes the best-fitting parameters and the process iterates again.

Essentially, the experimenter is an AI that is mathematically tuning the experiment in real time to maximize the probability of a conclusive finding. Science is awesome, isn’t it?

Leave a Comment

Computational Models Of No Mind

Imagine tens of thousands of years ago a couple of our ancestors manage to work out some symbolic representation for something, probably danger.

Symbols are, after all, mostly useful for social formation initially. An aggregate of communicating individuals will, in most situations, have a competitive advantage in spotting threats or food or whatever.

Eventually the symbols become more fine grained and they are used for a wider variety of social functions. I can signal that I want water or a rock or food, whatever.

Humans, eventually get the hang of recording the symbols in such a way that knowledge can aggregate. The already intense survival pressure on the capacity to manipulate symbols for oral communication goes up again.

This aggregation of symbols eventually becomes history and progresses to science.

I’ve been thinking about the Buddhist concept of “no mind” or not knowing. It’s such a counter-intuitive instruction since how does one use a brain, whose basic function is thought, to stop thinking.

Read the rest of this entry »

Leave a Comment

How will people be able to change?

One of the more realistic criticisms I have heard for the system I am planning is that people will find it difficult to escape the impressions of others.

Consider a study done in the UK. Researchers put the test subjects through a word-association test. The test was a ruse to allow them to expose the participants to a variety of words. They then kept track of how quickly the subjects walked from the laboratory to the elevator. Participants who had been primed with words associated with age and the elderly took a statistically significant thirteen seconds longer to make the trip.

Much as was suggested with the egg salad study, our brains are significantly more complex than we understand.

One of the things that will have to happen in order for this experiment to succeed is a change in the public perception and fear of technology. The narratives that we have in our media include:

  • Wall-E type where populations are lulled into into laziness when their every whim is met. The danger being that we don’t want to solve our problems because somehow humanity will grow bored.
  • 1984 types where the prevalence of invasive technologies have eliminated the capacity for creative action to the point that everyone’s a slave to the system.
  • The Matrix and Terminator are variations on that to where our technology grows beyond our capacity to control it and we have to fight it off.
  • I, Robot is an alternative version where well-meaning creators of technology make systems that, in their desire to protect humanity, limit its freedom.

If the next steps in our race’s development does involve technological components, then there will have to be a new narrative created. This interests me as I attempt to use technology for peace since peacemakers tend not to die of natural causes. I have no desire for a well meaning Christian to put a bullet in me while trying to stop the anti-Christ from putting the mark of the beast on everyone.

The narrative that needs to be come more dominant is a simply one. We create tools and we control them. Take the issue of priming as a case in point.

Read the rest of this entry »

Leave a Comment

What are the advantages of open research?

Open research has the potential to take much of an age old system and modernize it to take advantage of the amazing technical tools that we now have available. To borrow a phrase from open source software development, “many eyes makes problems shallow.” The academic process as it exists is verified by a small number of researchers at the outset and vetted by a expert review before publication.

This process works well in a world of limited communications capabilities, but that’s not the world we’ve lived in for the last five years at least. If we design the proper tools, the whole bit can be under continual review. When I am interested in more depth than I get in a paper, the entire process could be preserved online. If an error makes it into publication, it can be corrected in place and authoritatively rather than piecemeal as is the case now.

This is no simple task, but it is achievable.

An advantage that I am most interested in at this point in my academic career (a jaded graduate student) is that of pedigree.

Degrees are used because when bringing a new individual onto a project, be it at a University or for a corporation, it is important to know that this person is not a dumbass. A degree, from an Associates to a Ph.D., represents to interested parties that a student has learned certain knowledge, has a familiarity with particular set of practices, and is capable of some degree of critical thinking.

Imagine a system where students are able to contribute to active research projects of interest. They develop a pedigree based on their activities as well as their course instructions. Individuals contribute to projects that are of interest to them and in return they not only get the enjoyment of contributing to a project of interest, they are also developing a pragmatic certification of their skills.

The question of an individual’s qualifications can be answered with much finer granularity. This offers the hope of not only producing higher quality research but offering people the opportunity to focus on the elements of the process that they find enjoyable and are skilled at while leaving the other aspects to other people.

Leave a Comment

What is open research?

Open research is simply research performed in as open a manner as possible. From the generation of hypotheses, prostate to the experimental design, shop the data analysis and final publication, link the process is public and open to comment.

Peer review and vetting have a different meaning in this context because it is possible to view the entire process.

My approach to open research is based off of using the Ideapool network. Specifically:

  • Associating discrete and specific research contributions to a virtual identity.
  • Creating tools for the aggregate evaluation of research ideas and methods.
  • Creating tools for the analysis of structured datasets.

Leave a Comment

What do you hope to get out of TagItAll?

My goal with TagItAll is to produce two things:

  • Reasonably Reliable Virtual Identities — As people invest time into authenticating virtual identities and as those identities are enmeshed in networks of semantic associations, ailment there is an investment which can be added to be using the same identity in multiple contexts. This will enable trust in online systems in a way that is not possible with the level of anonymity that we have currently. The web has to be developed so as to control the actions of the least trustworthy whereas with invested identities it is possible to trust more thoroughly since anyone may add something to the pool to tarnish the identity.
  • Useful Semantic Network Data — This program, health if widely used, anabolics will generate a huge amount of useful statistical data about the meanings of words and their associations with different psychological characteristics. It will, at the least, permit the development of significantly more nuanced artificial intelligences for dealing with both personality and human language.

Leave a Comment

Are you qualified to do this research?

Nope. I know a bit about a lot of things, sale viagra but I definitely do not have the expertise to properly design and execute large scale research projects such as I plan on conducting in the Anarchist Grill.

I am unconcerned, not because I plan on somehow learning all that I need to know, but that I plan on attempting to convince more qualified researchers to do the work instead. If the plan works, I will have plenty to occupy my time. I figure with an interesting experimental setup, willing subject pool and available financing, there will be researchers interested doing the designing.

Leave a Comment

Why open a bar?

Why did I choose a bar as the business I would like to open to feed money into the profit-donation pool?

One of the issues that interests me in contemporary society is the growing divide between the political parties in American culture. I am interested in exploring is the role of the church and group formation relates to self-identification.

I think that it is a problem that we do not have many secular meeting places which focus on the development of community and the capacity from diverse groups to convene for entertainment. I have some experiments I would like to try in different types of activities to see how people respond and what entertains them.

An interactive bar will permit the collection of information and responsive feedback in whatever environmental characteristics can be set to be controlled by a computer. This will simultaneously permit the collection of large amounts of interesting data for generating computational modeling of social patterns and satisfaction.

I also want to create a bar simply because I think I can do a good job of it.

Imagine a bar that knows that you specifically are in it and will be tailoring the music, lights, volume and everything else to your liking.

Imagine a bar that if you dislike something you can access the AI running the environment and give direct feedback as to what you would like to see changed.

Imagine a bar that might invite you to a singles night specifically to introduce you to a special someone who, based on patterns in your data in the Ideapool, you have a high probability of compatibility with.

Imagine that same bar, after your meeting, adjusting its algorithms and learning to make better matches based on the opinions you insert in the pool after your meeting.

Imagine a bar where you know that if you go nuts and drop $100 on shots for your friends that all the excess of your money beyond what it takes to run the place will go into a pool for developing projects to improve the world.

Sounds like a pretty entertaining place, n’est pas?

Leave a Comment

What is the incentive to patronize profit-donation businesses?

Profit-donation businesses have a competitive advantage over a total profit business in that consumers will reinforce their self-concept of being involved in improving the world by choosing and intentionally self-limiting option.

In particular, businesses that choose to leverage the data in the Ideapool will be able to provide better services by mining and making available data.

For example, the business I would like to start is the Anarchist Grill, an interactive bar driven by a machine learning algorithm to maximize the contentment of the patrons.

Another idea would be a grocery store that inserts purchase records into the Ideapool. Those records could then be cross referenced with nutritional information as a part of an exercise tracker or aggregated as part of a budget or used to generate potential recipes. The public records could be used by producers to do market research.

People balk at the idea of having their privacy invaded, but they fail to realize that Wal-mart already knows everything that you’re buying. The day will come when Visa or someone else starts collaborating with major retailers to large scale data mining. There is simply too much profit to be had from exploiting this data. You can’t easily prevent the data from existing, but you can work to see that the benefit is to as many people as possible.

Leave a Comment

Older Posts »