Computer Engineering for artificial intelligence?

College Majors Engineering Majors
1

<p>I recently watched a documentary on artificial intelligence (called the Transcendent Man). It was narrated by Kurzweil and it made me aware of the growing importance of technology in this generation. I find the field interesting, however I haven't really ever done anything advanced concerning computers (have never coded or anything of the sort). This community has been very helpful in the past and seems to be quite informed on a multitude of subjects so I figured I'd come here first before deciding to take that path. I want to know what I'm getting into.
1. If I want to get into the field of artificial technology, should I major in computer engineering or computer science?
2. How hard is programming, and how much will I be doing in this field? Is it enjoyable to the people who do it? Or is it just a never-ending annoyance that comes with being in this field. Basically, what makes this field fun? </p>

<p>Sorry if it seems like I haven't done my research and am asking other people to baby me. I'm lazy, but if I really do need to do some research on my own, I'd appreciate it if you would point me in the right direction. </p>

<p>(you can ignore this last paragraph)
Last, I think I should provide a basis of what kind of person I am so that you can assess whether or not this is the field for me. (I am asking much of you collegeconfidential community. my bad.) I like science, biology especially. I love microbiology. I like it because everything makes sense; everything works. It's smart and therefore appeals to me. I think of myself as slightly higher than the crowd, and know that this view will be shattered when I go to college, which will be the College of New Jersey. I like to read about the brightest minds that humanity has seen, I like philosophy, I like to experiment with different drugs on a safe controlled basis in the pursuit of greater knowing (psibolin). I don't like authority. I like math, sometimes. I procrastinate though, my epic flaw. I over-analyze sometime, and am not always practical. I tend to overexaggerate things and get way too excited over things before actually doing much research. I'm a bit of an idealist, and tend to ignore the downfalls of my plans. Maybe I'm stupid, it's highly possible. Just gave you way too much personal info, which is ok with me considering this is anonymous. Feel free to troll me, make fun of me, anything. I will appreciate it. The crowd I may be attracting though may be able to take away my anonymity. That is the only problem I see with this thread. Done</p>

2

<p>If you watched Lord of the Rings, would you be studying to become an elf?</p>

3

<p>^ it’s perfectly fine for a movie to pique your interest in a field. I first became enamored of AI when I was 12 and saw AI: Artificial Intelligence with Haley Joel Osment. 10 years later, and I’m starting my PhD in AI next year. ;)</p>

<ol>
<li><p>It depends on what you want to do. I would recommend computer engineering only if you’re interested in robotics. That will have a greater emphasis on the EE aspect of the field. If you’re interested in the “software side” of AI (i.e. programming the brains, whether or not it’s in a robot), then CS is the way to go. Either way, you should be taking all the AI classes you can.</p></li>
<li><p>Programming isn’t hard. As anyone who’s programmed before will tell you, once you learn your first language, the rest are easy. Some people love to program, others think it’s boring and just a tool for research. Not a huge portion actually dislike it; at worst it’s a little tedious. I personally like to program - writing lines of code that, when run, actually do something, in this case make the computer more intelligent, is fascinating to me. Don’t worry that you’ve never coded before; plenty haven’t when they start out in AI. When you do learn to program, you’ll find, like everyone else has, that it’s pretty anticlimactic (i.e. not painful) but still awesome.</p></li>
</ol>

<p>As for whether you’d like AI, it’s hard to say. Biology itself isn’t terribly related, unless you are planning on doing computational neuroscience or bioinformatics (both of which draw on AI). The reason you like biology is the same reason I like AI: “everything makes sense; everything works.” This is especially true in AI, which endeavors to formalize the human mind along computational, statistical, logical lines. The formal sciences - math, logic, statistics, CS - are the “inherently true” sciences, the ones whose truths exist independent of external discoveries. AI draws on the formal sciences in an attempt to recreate the mind, with the theory being that the mind is nothing but a computing machine whose intricacies we need to uncover in order to model it on a computer (the cognitive assumption). AI also draws on linguistics, psychology, neuroscience, and philosophy. Do these areas, in addition the formal sciences, appeal to you? You can of course avoid certain areas depending on which subfield of AI you focus in. For example, I don’t do any work in image processing or computer vision because it’s not my interest.</p>

<p>The best way right now for you to find out whether you like AI is to just read the Wikipedia article on AI. It lays out the field pretty well. Pay special attention to the “problems” section, which will give you an idea of what you’d be working on if you pursued AI. If you decide you like it, I’d recommend getting Artificial Intelligence: A Modern Approach by Russell & Norvig. It’s the standard AI book and can even be downloaded for free in ebook format (google it).</p>

<p>By the way, if you liked Transcendent Man, you should watch The Singularity Is Near. It’s the other Kurzweil film but doesn’t focus on him. Take it with a grain of salt, but the movie is good for getting you to think about the fundamental/philosophical issues in AI.</p>

4

<p>Awesome information. Thank you so much. and, boo to you James Madison, you heroin dealing scum. anyway, based on your response, I think that the software part of computer engineering and AI would be perfect for me. As of now, I don’t see a reason why I wouldn’t like the field. My interdisciplinary study (kind of like a minor) is cognitive science, which involves anthropology, biology, AI, linguistics, philosophy, psychology and probably something else that I am forgetting. I plan on reading the article after work and probably buying/downloading the book you provided. All that’s left to do is call my school and figure out which courses I should be taking in order to get the best AI education possible. Are there any schools that are particularly good in AI, that I should be looking into. If you know about TCNJ at all, let me know what you think I should do. Currently, my plan is to get very good in the AI field and help contribute something to what Kurzweil talked about in the Transcendent Man, how possible do you think that is. Also, assuming I do go on to get my pHd, what’s the salary looking like? Last, how competitive is this field? Like I said, I think I am above the mark for the most part, but maybe I’m getting ahead of myself. I have 5’s on a.p. bio and english without doing much work (not bragging) and 4’s on us history and psych, which could have been 5’s imo. I got a 2020 SAT without studying or taking any classes, without really doing any work in highschool (which I do regret but here I am again giving you more of a basis of what I am). So, what do you think</p>

5

<p>Cognitive science is the heart of AI; cogsci is essentially AI, though it can be applied to other fields as well, not just AI.</p>

<p>

</p>

<p>Yes, there are many schools that are excellent in AI. Here’s a good list - yes, it’s a ranking, but it does manage to name most of the top AI schools:</p>

<p>[Best</a> Artificial Intelligence Programs | Top Computer Science Schools | US News Best Graduate Schools](<a href=“http://grad-schools.usnews.rankingsandreviews.com/best-graduate-schools/top-science-schools/artificial-intelligence-rankings]Best”>http://grad-schools.usnews.rankingsandreviews.com/best-graduate-schools/top-science-schools/artificial-intelligence-rankings)</p>

<p>Yes, it’s very possible - even probable - for you to contribute something significant to AI. What’s great about the field is that while it’s been around for 60 years (so you have some foundation), its methods are constantly changing. Before the 1990s, most work in AI was on the symbolic/rule-based/logical side. But with the increase in power of computers, we’ve been able to apply statistical techniques and make great strides. Since we’ve done this, a number of technologies have been commercialized:</p>

<p>

</p>

<p>At the same time, we need even more computational power to make systems that are both accurate (i.e. intelligent in a broad sense) and robust (i.e. scalable to the size of, say, the Web). For many years that’s been the primary tradeoff, but if Moore’s Law continues - and even if it doesn’t - we’ll have computers with CPUs and memory capable of processing in reasonable time enormous amounts of information. And doing it accurately.</p>

<p>For example, you might look into memristors, which many think will be the future of computing memory and is especially relevant to AI since it’s inherently neuromorphic. That’s a new technology that has promise for the future and that will likely shake up AI. The point is, there are continually developing technologies that change the landscape for AI, such that there is no better time to get involved in AI than now. It is now that tons of money is being dumped into AI research; now that tons of companies are incorporating AI into their technologies; and now that we need AI more than ever. In fact, given that we’re in the information age - and the amount of digital information is increasing to absurd heights (see [this](<a href=“Byte - Wikipedia”>Byte - Wikipedia)</a>) - we have no hope in the future for sifting through data without intelligent machines. It really *is *the future, and so now is the best time to get involved in it and make a significant, game-changing contribution.</p>

<p>As for salary, those with PhDs in AI do really well - I don’t know what the average is, but I’ve heard at many schools it can be $150k starting. AI researchers are in high demand, and still there isn’t enough supply to sate that demand. So the competition itself isn’t particularly strong, as in other older, more-established disciplines. Another great thing about AI is that since it crosses so many disciplines, coming up with a new idea for AI isn’t particularly hard. You could take one idea from psychology and a method from statistics to come up with a new machine learning algorithm that ends up classifying object data much better than other methods. Most of it is uncharted territory, even though researchers have been working on it for decades. That territory is so broad that the competition isn’t particularly stiff (though that isn’t to say there aren’t crappy ideas and good ideas, or that people in the field can’t tell one from the other).</p>

<p>I don’t know about TCNJ itself, but a glance at their courses tells me you will be able to major in computer science or something related, and take AI courses in addition to the cognitive science courses (which are helpful in AI). I assume, also, that you’d be able to do research with a professor - the CS site lists a research mentorship program. Cogsci might also have that. So if you do well in classes, do research, and get involved with faculty (through research) for recommendations, you’ll have a *very *good shot at getting into the top AI schools like Stanford, MIT, CMU, etc. for PhD, even master’s.</p>

6

<p>I’m so excited to get started. It’s what I’ve been looking for for a while now. What laptop would you recommend getting? Any in particular? And, what’s your opinion on the whole ‘living forever by 2049’ thing… is it possible?</p>

7

<p>Kurzweil, lol.</p>

<p>Here’s what some folks in the science world have to say about Kurzweil’s predictions:</p>

<p>Biology blog Pharyngula, run by PZ Meyers:
[Singularly</a> silly singularity : Pharyngula](<a href=“http://scienceblogs.com/pharyngula/2009/02/singularly_silly_singularity.php]Singularly”>http://scienceblogs.com/pharyngula/2009/02/singularly_silly_singularity.php)</p>

<p>Well respected, Pulitzer prize winning author and acclaimed cognitive scientist, Doug Hofstadter (author of Godel, Escher, and Bach):
[Alan</a> Fineberg: Smack talkin’ D.R. Hofstadter delivers sick iceburn on R. Kurzweil](<a href=“http://alanfineberg.blogspot.com/2009/07/smack-talkin-hofstadter-lays-smackdown.html]Alan”>Alan Fineberg: Smack talkin' D.R. Hofstadter delivers sick iceburn on R. Kurzweil)</p>

<p>Original interview where Doug lays into Kurzweil’s nonsense here:
[An</a> Interview with Douglas R. Hofstadter, following ‘‘I am a Strange Loop’’](<a href=“An Interview with Douglas R. Hofstadter, following ''I am a Strange Loop''”>An Interview with Douglas R. Hofstadter, following ''I am a Strange Loop'')</p>

<p>Awesome Newsweek article by Dan Lyons (FakeSteveJobs, lol:
[Ray</a> Kurzweil Wants to Be a Robot - Newsweek](<a href=“http://www.newsweek.com/2009/05/16/i-robot.html]Ray”>Ray Kurzweil Wants to Be a Robot)</p>

<p>My issue with Kurzweil is that he’s basically trying to predict something that’s entirely unpredictable. Forget the fact that we can’t even accurately predict the outcome of seemingly simple systems when we try to model them (the weather, patterns of swirling smoke or fog, any system that exhibits a sensitive dependence to initial conditions) - he’s trying to predict the future OF ALL TECHNOLOGY. The thing about paradigm shifts, though, is that they usually come from unexpected discoveries - unforeseen events. Also, his entire living forever by 2049 is … I don’t even know. When I think about the various interactions taking place within a single cell, and the all of the things that we still have no idea are happening in regards to protein transcription and structure and function - I just feel overwhelmed by how much I don’t know. I doubt we will be living forever by 2049. I hope I’m wrong, but it seems unlikely.</p>

<p>P.S. I am NOT telling you to not go into AI research. Programming is awesome! It’s a lot of fun, logically challenging, and you’ll find joy/tears/etc… within the realm of mathematics and computation. I particularly find computer vision to be fascinating (I build robots for fun, but they are pretty crappy right now, because improving their response to stimulus requires more programming knowledge than I have - it’s tough to make a robot respond ‘intelligently’ to objects).</p>

<p>I’m just saying that I think Kurzweil is kind of a hack and a snake oil salesman. But if he gets more people interested in the sciences, then that’s great. Once you learn more about this stuff, you’ll see why his claims are ridiculous anyway.</p>

8

<p>Additional commentary on Kurzweil’s belief that it will only take 1 million lines of code to simulate the human brain:</p>

<p>[Ray</a> Kurzweil does not understand the brain : Pharyngula](<a href=“http://scienceblogs.com/pharyngula/2010/08/ray_kurzweil_does_not_understa.php]Ray”>http://scienceblogs.com/pharyngula/2010/08/ray_kurzweil_does_not_understa.php)</p>

<p>Man, I have a lot of respect for this guy as an inventor and computer scientist, but woah. I mean, whatevs. Newton had strange/esoteric beliefs as well, but he tried to keep them under wraps. Kurzweil is so open with his nuttiness. Anyway, here’s a fun site that hosts long term bets:</p>

<p>[02002-02029</a> (27 years): By 2029 no computer - or “machine intelligence” - will have passed the Turing Test. - Long Bets](<a href=“http://longbets.org/1/]02002-02029”>02002-02029 (27 years): By 2029 no computer - or "machine intelligence" - will have passed the Turing Test. - Long Bets)</p>

<p>That one is between Kurzweil and Mitch Kapor of Lotus:</p>

<p>“By 2029 no computer - or “machine intelligence” - will have passed the Turing Test.”</p>

<p>We’ll just have to see how it goes!</p>

9

<p>Kurzweil… he and Wolfram should be thrown in the same loony bin.</p>

10

<p>I really wish there was more time in a day… and less work to do in each… and more money to be had in doing so. I’ll read those articles eventually, but I am saddened that it will probably kill my hopes of living forever. I still intend on going into artificial intelligence. Again, what laptops should I be looking at, and how was your school experience (to those who took the computer science path of artificial intelligence)?</p>

11

<p>@Phantasmagoric: Your first reply had my eyes watering (from joy)… </p>

<p>Haven’t looked (or not long enough) for a free copy of ‘Artificial Intelligence: A Modern Approach’ (or maybe I have it on my machine at home already and have forgotten about it).
Watching ‘Transcendent Man’ & ‘The Singularity is near’ as soon as possible…</p>

<p>Thanks a lot…</p>

12

<p>AI is generally part of computer science, however, it can also be viewed on it’s own from a more philosophical view. </p>

<p>In some schools AI is part of cognitive science. If you look at the curriculum for Cognitive Science at Johns Hopkins, you can see how different areas relate to AI.</p>

13

<p>I’m really interested in AI and robotics at the same time… Which is better for me to do?? [Comp Science OR Comp Engg]</p>

14

<p>that was a long bump.</p>

<p>ashwinlokkur, probably computer science</p>

<p>to the naysayers,</p>

<p>I don’t care if you believe in Kurzweil’s predictions or not</p>

<p>You know that information technology grows at an exponential rate</p>

<p>So it’s not inconceivable that in a few decades at most, we will have enough technology for a machine to pass the Turing test</p>

<p>And once that happens, things get very, very interesting, very quickly</p>

15

<p>Please reply as soon as possible.I am confused as to whether I should take Comp Science or Comp Engineering . I want to concentrate on Artificial Intelligence as a specialization.Out of CmpE and CS , which would provide me a better start?</p>

16

<p>Phantasmagoric - You seem to know a lot about it.So please reply to my post</p>

17

<p>AI is more in the realm of CS than CE. Take a look at the course offerings at the top schools and see what department they are in…</p>

18

<p>I’m a Computer Science grad student who’s focus is on machine learning and vision (both of which are subsets of A.I.). IMO, backgrounds in CS and CE are both equally appropriate for a focus in A.I. However, I think a background in mathematics (with a focus in computer science) is more ideal for an A.I. researcher.</p>

<p>In short – A.I. is a lot of math.</p>