Algorithms In Psychology

[AlphaRXX8]

Does any experts here think that it is possible (it should be) to write a set of algorithms to sequence human behaviour? As founder of Behaviourism, J.B Watson, seems to think that human behaviour can be quite mechanical.

 

As the quote goes: "Give me a baby, and I can make him into any man you want."

The experiment on ethnicities and race concerning mentality was not carried out, he died : (

 

I'm sure we all know that the credibility of Psychology and its brothers, can be improved a lot more, as currently, it is entirely based on probability and likelihood (increases as the number of legitimate studies correlates), however, it is not as credible as other sciences.

 

I currently finished year 11, so there's nothing I can do to improve the science until I reach university.

[EdEarl]

I am not an expert, but just read (minutes ago) an article that may help answer your question.

 

http://www.mediadesk.uzh.ch/articles/2013/problemloesen-lenkt-die-verarbeitung-von-sinneseindruecken_en.html

 

Many researchers in AGI and neurology believe simulating the brain is possible with computers and programs. Some have predicted the Singularity within this century. I believe

has made very important discoveries and developed advanced AGI software necessary for the Singularity. Although, I believe the Singularity will not be a single technology and event, rather several. For example there may be intelligent computers both with and without emotions.

 

Already AI plays chess, Jeopardy, and drives cars. People are still better at driving cars, but it will not be long before an AI driver is safer than a human. How much smarter does AI need to be an example of the Singularity? I know the military and businesses are very interested in smart machines, and they have money to invest. Advancements are coming fast, and will come faster.

[AlphaRXX8]

No, I'm not talking about AIs, I'm talking about the use of mathematics in the realm of Psychology, to increase its credibility, and to be used to predict human behaviour and actions with 100% accuracy.

[EdEarl]

Sorry, IDK.

[John]

I don't think we understand the workings of the human mind well enough yet to really answer that question, but you might enjoy reading about mathematical psychology. There is also the idea of the quantum mind to consider, though I don't think it's really a mainstream hypothesis. Also, if it turns out that we can create an artificial intelligence as capable as a human brain, then perhaps we'll be able to run simulations quickly enough to gather much more data than would be possible in human experiments, thereby improving our ability to predict how human subjects will act in various situations.

 

However, even if we show that classical mechanics can be used to predict human behavior, I wouldn't expect 100% accuracy to be attainable (though I'm by no means an expert). We might get close, but perfection is a tall order.

[Ophiolite]

No, I'm not talking about AIs, I'm talking about the use of mathematics in the realm of Psychology, to increase its credibility, and to be used to predict human behaviour and actions with 100% accuracy.

Since the AIs would function through the programming of algorithms, I would have thought that you very much are talking about AIs. I would be interested in knowing why you think you are not.

[Popcorn Sutton]

I think that we can get a good grasp of behavior psychologically using mathematics, but it's going to take the right algorithm. Simply put, you need to know about the workings of the brain and of knowledge. It's not a single discipline that goes into that effort. Multiple fields are required and you need to have a firm grasp of theoretical concepts from fields like neuroscience, physics, computer science, mathematics, etc. I would think that the best mathematical tool to use for this type of study is statistics, and a hint of elementary algebra. I would think that one of the main hypotheses involved in this type of research is pattern detection, and I would also think that you would need to keep the behaviorist dogmas in mind while doing so. Conditioning is just a given, but you have to know exactly how it happens neurologically, which I think has a correlation with P600's, or positively charged electrical activity. Experiments have been done on people reading what are called "garden path sentences", where you have a parsing problem reading the sentence, and a P600 occurs when the parsing problem happens, but it can't be repeated with the same sentence. An example of one of these sentences is "the horse raced past the barn fell."

 

I think that by using mathematics, we can gain insight into psychological disorders, but, to me, it seems more of an issue involving pragmatics than psychology.

Edited July 27, 2013 by Popcorn Sutton

[EdEarl]

There are mathematical models of neurons, but they are not completely accurate since things are unknown about neurons. Models of complex system invariably suffer from the model being incomplete and from being too complex to solve or to solve in real time.

 

The n-body problem is solved for 2 bodies, and for special cases of 3 bodies, but cannot be solved for 4 or more bodies. Thus, numerical methods (I.e., simulations) are needed for 4 or more bodies. This problem is much simpler than the phychology of a person; thus, it seems unlikely a good psychological model exists.