AI or HCI, That is the Question.

[Christina Brown]

Hi everyone. I am a new TheGradCafe member. I am very excited about the graduate school process. I am planning to apply this fall. Right now I am doing intensive research on Computer Science and I torn between AI and HCI as possible specializations. I know that in Stanford, I can possibly do a dual specialization in these two sub fields but I am torn between the two because they will both contribute to the extensive technology and mobile dialogue in the next 10-25 years.

Are there any other graduate schools out there that allows students to either specialize in both fields or take supplementary classes in one field and specialize in another? I am leaning towards HCI as my primary field of study since I have a B.A in sociology and I am passionate about design and produce development. I don't know if AI would be a better choice, since it is more math based. I am really good in math despite my liberal arts background. What are your insights in these fields of study and the graduate school process? Thank you for your time!

[stopcallinmesqrlboy]

You could study HCI as it is related with AI. Also, AI is a huge area-- what specifically in AI interests you?

You say you are good in math, but as a sociology major, do you have experience with logic, probability, statistics and/or linear algebra? These are important areas of math that are often used in AI research.

[adelashk]

Hi everyone. I am a new TheGradCafe member. I am very excited about the graduate school process. I am planning to apply this fall. Right now I am doing intensive research on Computer Science and I torn between AI and HCI as possible specializations. I know that in Stanford, I can possibly do a dual specialization in these two sub fields but I am torn between the two because they will both contribute to the extensive technology and mobile dialogue in the next 10-25 years.

Are there any other graduate schools out there that allows students to either specialize in both fields or take supplementary classes in one field and specialize in another? I am leaning towards HCI as my primary field of study since I have a B.A in sociology and I am passionate about design and produce development. I don't know if AI would be a better choice, since it is more math based. I am really good in math despite my liberal arts background. What are your insights in these fields of study and the graduate school process? Thank you for your time!

Hi Christina,

Welcome to TGC. Part of what the adcomm wants to see is that you are as specific as you are since graduate school is more about depth as opposed to breadth. Instead of trying to dual specialize, focus on one area.. AI is a relatively more theoretical field with many people who specialize in it aim for the academic path as opposed to going back to the industry. So you should ask yourself, are you interested in a theoretical field that requires a lot of math and logic?... I am not very familiar with HCI but I do know for some schools HCI does not fall under the CS department.

Also, you mentioned that you are interested in HCI due to your bachelor's background but what interests you in AI?

Edited April 12, 2011 by adelashk

[Christina Brown]

You could study HCI as it is related with AI. Also, AI is a huge area-- what specifically in AI interests you?

You say you are good in math, but as a sociology major, do you have experience with logic, probability, statistics and/or linear algebra? These are important areas of math that are often used in AI research.

I have taken classes in statistics and algebra in college and in high school. Machine Learning and Robotics so far interest me but I need to do more research before I can make my decision on what field of CS I want to pursue.

Hi Christina,

Welcome to TGC. Part of what the adcomm wants to see is that you are as specific as you are since graduate school is more about depth as opposed to breadth. Instead of trying to dual specialize, focus on one area.. AI is a relatively more theoretical field with many people who specialize in it aim for the academic path as opposed to going back to the industry. So you should ask yourself, are you interested in a theoretical field that requires a lot of math and logic?... I am not very familiar with HCI but I do know for some schools HCI does not fall under the CS department.

Also, you mentioned that you are interested in HCI due to your bachelor's background but what interests you in AI?

Yea, I think its my liberal arts background that is kicking me in the butt right now ><. What interests about AI is how mobile devices and search engines are using AI to determine what we want based on our location in a region of the world or what our initial interests are for example. After graduate school, I plan to work in industry so AI may not be a good choice for me as you pointed out. I also heard stories of people switching specializations in a CS program for their Masters after getting in as well. I will keep my liberal arts background in check though.

Edited April 12, 2011 by Christina Brown

[beanbagchairs]

Hi Christina,

Welcome to the forum!

That sounded like HCI to me.

[giantfox]

George is right. AI and HCI have many overlapping applications. Look at aIED, ITS, and the IUI conferences for some of the subfields where these two ideas meet heavily. I had the same dilemma about a year ago, so good luck to you!

[adelashk]

You are missing the point here, AI and HCI are somewhat related but they are different fields. I knew a lot of HCI graduate students back in my alma mater (top 10 in engineering) who were very interested in AI and, against my advice, decided to take an AI class in the EECS department. They ended up dropping the class within 2 weeks never to venture into AI again. As you said, you need to do more research into the AI field and find out exactly what interests you about it. It is not all about cool robotics projects and shiny applications.

It will not give you an advantage if you talk about doing HCI, AI, and some other field.. Be specific. If you are applying to top CS schools without a CS degree, consider taking the CS Subject GRE.

[csperson]

You are missing the point here, AI and HCI are somewhat related but they are different fields. I knew a lot of HCI graduate students back in my alma mater (top 10 in engineering) who were very interested in AI and, against my advice, decided to take an AI class in the EECS department. They ended up dropping the class within 2 weeks never to venture into AI again. As you said, you need to do more research into the AI field and find out exactly what interests you about it. It is not all about cool robotics projects and shiny applications.

It will not give you an advantage if you talk about doing HCI, AI, and some other field.. Be specific. If you are applying to top CS schools without a CS degree, consider taking the CS Subject GRE.

I agree, these two are different fields

[giantfox]

I agree that they are separate. Even in the cases where they overlap, these projects have multiple students working on the separate aspects of these problems. There will be a graduate student who is familiar with HCI practices doing the evaluation as well as an AI grad student working on the implementation of the program, who may base their implementation from information derived from a previous HCI evaluation. If your goal is to work as a PI on a project involving the two it is probably better to be well versed in AI, whereas HCI concentration will probably focus more on understanding people. But this is hard to say, since HCI is such a new field. Most of the PI's I've seen working on such projects have degrees in AI, but that might change now that there are much more HCI based programs out there.

[Christina Brown]

You are missing the point here, AI and HCI are somewhat related but they are different fields. I knew a lot of HCI graduate students back in my alma mater (top 10 in engineering) who were very interested in AI and, against my advice, decided to take an AI class in the EECS department. They ended up dropping the class within 2 weeks never to venture into AI again. As you said, you need to do more research into the AI field and find out exactly what interests you about it. It is not all about cool robotics projects and shiny applications.

It will not give you an advantage if you talk about doing HCI, AI, and some other field.. Be specific. If you are applying to top CS schools without a CS degree, consider taking the CS Subject GRE.

Even if the graduate school does not require a CS GRE Test? And you are probably right about AI. I'm going to research a little bit more on the field before I narrow down my schools. Thanks!

Edited April 12, 2011 by Christina Brown

[Christina Brown]

George is right. AI and HCI have many overlapping applications. Look at aIED, ITS, and the IUI conferences for some of the subfields where these two ideas meet heavily. I had the same dilemma about a year ago, so good luck to you!

Could you tell more about your situation a year ago? I love stories.

[adelashk]

Even if the graduate school does not require a CS GRE Test? And you are probably right about AI. I'm going to research a little bit more on the field before I narrow down my schools. Thanks!

A lot of schools do not require a CS GRE but some recommend it. If your undergraduate major is not related to CS and you have not taken CS classes then top grad schools have no way of gauging your potential as a CS student/researcher. You might be a god in Sociology but that only reflects that you're a great social science student. The competition among candidates applying to top schools is not based on who is qualified or not, almost all if not all are qualified. If you take the CS GRE and do well then you will have a fighting chance at an admit. My advice only applies to top grad schools and undergrads majoring in a field not related to CS. For example, many physics/math majors get into top CS schools as compared to psychology/economics/etc.. majors.

[BKMD]

Are you planning to do a MS or PhD? Since your undergrad is a non-technical major, departments may want you to do an MS before going on to a PhD. If you do a higher degree in CS, then you'll be expected to already know core things like programming, data structures, algorithmic analysis. You might need to do some catching up on background material. The Subject GRE is recommended if your undergrad degree is in a different field, since otherwise the schools do not know how well you understand the background material. However, this test is pretty tough (it can take a lot of studying even for a CS major). It's also possible that if you are applying directly into an HCI program (I know CMU has this) then they might not require as much of a CS background.

You may also consider directly applying to a social science department rather than CS, since often those departments still do computational research. It is actually fairly common for AI researchers to be in linguistics or cognitive science departments rather than computer science, if that is the area they are stronger in (in other words, if your BS is in linguistics and you want to do language processing, you might still do your PhD in linguistics even though you'll be doing computational stuff - the NLP group at Stanford is pretty evenly split between ling. and CS students, for example). Computational social science is becoming a popular sub-area so you might fit well into that - for example there is a center for that at UMass (http://www.cssi.umass.edu/). I think there are a lot of cool ways you can combine your background in sociology with computer science and statistics, so that might be something to consider.

Anyway, I agree with others that you should be specific about what you want to do when you apply - but you still have several months to decide. It sounds like you already have good ideas about what you find interesting and what you'd like to do, so just keep thinking about it. Maybe read some textbooks for AI and HCI (the standard text for AI is Russel&Norvig; I don't know much about HCI) to get a sense of how they compare. Take a look at research papers in the different fields and see what excites you more.

Good luck!

[Christina Brown]

Are you planning to do a MS or PhD? Since your undergrad is a non-technical major, departments may want you to do an MS before going on to a PhD. If you do a higher degree in CS, then you'll be expected to already know core things like programming, data structures, algorithmic analysis. You might need to do some catching up on background material. The Subject GRE is recommended if your undergrad degree is in a different field, since otherwise the schools do not know how well you understand the background material. However, this test is pretty tough (it can take a lot of studying even for a CS major). It's also possible that if you are applying directly into an HCI program (I know CMU has this) then they might not require as much of a CS background.

You may also consider directly applying to a social science department rather than CS, since often those departments still do computational research. It is actually fairly common for AI researchers to be in linguistics or cognitive science departments rather than computer science, if that is the area they are stronger in (in other words, if your BS is in linguistics and you want to do language processing, you might still do your PhD in linguistics even though you'll be doing computational stuff - the NLP group at Stanford is pretty evenly split between ling. and CS students, for example). Computational social science is becoming a popular sub-area so you might fit well into that - for example there is a center for that at UMass (http://www.cssi.umass.edu/). I think there are a lot of cool ways you can combine your background in sociology with computer science and statistics, so that might be something to consider.

Anyway, I agree with others that you should be specific about what you want to do when you apply - but you still have several months to decide. It sounds like you already have good ideas about what you find interesting and what you'd like to do, so just keep thinking about it. Maybe read some textbooks for AI and HCI (the standard text for AI is Russel&Norvig; I don't know much about HCI) to get a sense of how they compare. Take a look at research papers in the different fields and see what excites you more.

Good luck!

Thanks for the feedback. I am not planning on getting my Ph.D, just my Master's. As I narrow down my list, I will keep the subject test in mind and find out if it is recommended for non-CS majors.

[scyrus]

Thanks for the feedback. I am not planning on getting my Ph.D, just my Master's. As I narrow down my list, I will keep the subject test in mind and find out if it is recommended for non-CS majors.

It almost always is. It would be good for you to take it, anyway, because it will allow you to learn a lot about CS in the process.

And yes, AI and HCI have some overlap but their focus is completely different. Ask yourself if you want to spend your time learning how to make computers understand humans better (which seems to be the part of AI you are interested in) or whether you want to use those techniques to improve human experiences with computers, because there is a difference. The former will require a firm grounding in logic, machine learning and knowledge-based artificial intelligence where as the latter will require perspicuity in finding applications where applying those theoretical techniques will improve the status quo (sometimes in significantly better, startling ways).

This is, of course, not to say that people in HCI can't invent new techniques for tackling problems and people in AI can't apply their techniques practically. But as an incoming Master's students with no prior CS background, you will not have too much time to spread your attention.

This is just my opinion as an undergraduate who is specializing in AI and who is entering a HCI PhD program in the fall.

[DorkRawk]

I'm in a somewhat similar position as you. I was a CS undergrad and then did my masters in HCI. While I was taking a cog sci course for my HCI degree I became interested in AI and started trying to take more grad level classes related to that field. I recently had an unsuccessful round of PhD applications where I had mentioned an interest in studying AI (though I wasn't specific enough with what I really wanted to study) and a secondary interest in HCI. AI and HCI are not the same thing and they are both WAY too broad to say you want to "study AI" or "study HCI".

I think it's ok to have interests in both, but make sure that you have SPECIFIC interests in both and that you can speak intelligently about those specific areas of interest. There certainly are areas of overlap and if that's what you're interested in talk about specific work in that area.

[Christina Brown]

I'm in a somewhat similar position as you. I was a CS undergrad and then did my masters in HCI. While I was taking a cog sci course for my HCI degree I became interested in AI and started trying to take more grad level classes related to that field. I recently had an unsuccessful round of PhD applications where I had mentioned an interest in studying AI (though I wasn't specific enough with what I really wanted to study) and a secondary interest in HCI. AI and HCI are not the same thing and they are both WAY too broad to say you want to "study AI" or "study HCI".

I think it's ok to have interests in both, but make sure that you have SPECIFIC interests in both and that you can speak intelligently about those specific areas of interest. There certainly are areas of overlap and if that's what you're interested in talk about specific work in that area.

Im going with HCI. Can't wait until this fall!

[starmaker]

I'm a bit late to this thread, but for the record, I think AI and working in industry can go together (I'm in AI and working in industry, so obviously I have some bias here). Government contractors and EECS-oriented FFRDCs take AI people. Robotics companies take AI people. Increasingly, biotech/bioinformatics companies take AI people (specifically, machine learning/data mining people). Any company working on information retrieval or Internet search takes AI people. I've been through two (successful) job searches in the last year; I've seen who takes AI people.

However, as has already been pointed out, AI is huge. Theoretical AI has fewer industry applications than the application of machine learning to information retrieval, for example. I had mostly worked in machine learning and sensor processing in the past (and enjoyed them), but now I am working in morphogenetic robotics, which is a lot of fun and closer to what I would want to do in PhD work.

[Christina Brown]

I'm a bit late to this thread, but for the record, I think AI and working in industry can go together (I'm in AI and working in industry, so obviously I have some bias here). Government contractors and EECS-oriented FFRDCs take AI people. Robotics companies take AI people. Increasingly, biotech/bioinformatics companies take AI people (specifically, machine learning/data mining people). Any company working on information retrieval or Internet search takes AI people. I've been through two (successful) job searches in the last year; I've seen who takes AI people.

However, as has already been pointed out, AI is huge. Theoretical AI has fewer industry applications than the application of machine learning to information retrieval, for example. I had mostly worked in machine learning and sensor processing in the past (and enjoyed them), but now I am working in morphogenetic robotics, which is a lot of fun and closer to what I would want to do in PhD work.

Human-Robot Interaction (HRI) I hear is also popular and blends with HCI and other disciplines. I think I found my sandbox playmate.

Edited April 19, 2011 by Christina Brown

[qqyyzz]

I have taken classes in statistics and algebra in college and in high school. Machine Learning and Robotics so far interest me but I need to do more research before I can make my decision on what field of CS I want to pursue.

Yea, I think its my liberal arts background that is kicking me in the butt right now ><. What interests about AI is how mobile devices and search engines are using AI to determine what we want based on our location in a region of the world or what our initial interests are for example. After graduate school, I plan to work in industry so AI may not be a good choice for me as you pointed out. I also heard stories of people switching specializations in a CS program for their Masters after getting in as well. I will keep my liberal arts background in check though.

Hi Christina, I can't speak for HCI but if you're interested in the machine learning side of AI you will need to take a fair amount of mathematics. You mentioned you had taken statistics, but for a graduate degree in this area they would probably expect you to have a calculus based probability/stats class and probably a discrete math course. Also, you mentioned algebra in college/high school, was this course in elementary algebra (variables, arithmetic, polynomials), linear algebra (matrices, vector spaces, linear transformations) or abstract algebra (groups, rings)? You will need to be comfortable in linear algebra; the concept of vector spaces is used in many places and many things will be written in matrix form.

Don't worry though, if you don't have these things, you can always take a few courses before you apply. I was an econ major and subsequently found out I was more interested in data mining/large data set problems and have decided to take a bunch of math courses so that I am ready to apply next fall for stats programs.