hey, i`m at uni just now and doing AI as a side subject and were meant to write an essay for the 19th of december. the thing is that i dont know what to write about. these are the essays other people are doing in my class: Concepts of evolution theory in Artificial Intelligence Development of an Intelligence theory to solve AI AI in Computer Security AI in Computer Games Semantic web Computer Vision in Surveillance Under The Skin Genetic Algorithms/Planning with AI Representations of AI in Film Achievements of AI in everyday life Autonomous Vehicles AI, Is it "real" intelligence? How the media have influenced AI. AI and its use in Forensics The constant redefinition of the term AI Intelligent Filtering AI In Contemporary Art AI development projects and uses within the military i was looking for some advice because i really dont know what to write about. i`m open to any suggestions Thanks Thierry
Computation and emotion - comparisons of the human brain to machines. There was a substantial essay question on that in our computational theory exam in final year. Should give you much to talk about
I'm doing a masters in AI at the moment, and my favourite area is evolutionary algorithms. There's a ton of stuff to talk about. If you want some more information, PM me. AH
there's a difference between true AI as defined in academia and a rule-driven state machine granted i don't actually know how the AI director in l4d works, but you can get very very good results from static rules executed in strict order - it leads to something known as emergent behaviour, and it's pretty much the only thing you've seen in games so far. There are a few games that are beginning to use honest-to-goodness AI, Black&White being one of the first real poster children (yeah it was buggy as hell). The dog in fable 2 is pretty AI-driven too, supposedly. Molyneaux said they used a lot of the creature code in there.
I did my final year project on "AI Character Recognition using applied Artificial Neural Networks". There is a lot of in dept research, theory and maths involved so should make a good essay.
Thats what i was going to suggest, evolutionary/genetic algorithms. An absolutely fascinating subject in my opinion!
This may have already been covered in your course, but one of the biggest things to consider with AI is what we actually define as intelligence. With computers, we can look back at something like the concept of a Turing Machine as being a pretty decent defining line for whether something classifies as a computer, but to my knowledge, nothing in the field of AI equates to this at all.
How about an essay entitled "A.I. And how I intend to use it to take over the world mahahhahahhahaha." evil laugh optional of course.
Not important IMO. What we can do using AI and what techniques we can use are much more important than how we define the boundaries.
If you cannot define what AI is, how on earth do you propose to claim you are using it? Or are we all to reduce our command of the english language to that of marketing people?
Defining AI is something of a roundabout, because of this it's often best to ignore the definition of AI, as long as you are able to understand AI and provide examples of it. There are countless debates about AI because one argument is, does AI exist? Is it really intelligence or just a whole load of alogrithms and programming? Answering this question will get you nowhere so there is no point trying to define it, just try to understand it. On topic, I'd go for writing an essay on Evoluntionary Alogirthms and Genetic Programming, it's pretty easy to make examples of it and there is plenty to write about (including flaws and benefits of each Evolutionary Alogrithms), I did it and enjoyed it
Possibly, but emergent behaviour is observable in anything that gets sufficiently complex. Our own consciousness is a product of emergence. The problem with AI is that we really struggle to define "Intelligence" as it is (although it is easier to spot its absence... ). So how can we define "Artificial Intelligence"? Another aspect you can explore is whether AI can ever be the same form of intelligence as our own, or something fundamentally different. Our intelligence is a product of the environmental and survival demands imposed upon us as organic lifeforms (food, warmth/shelter, safety, procreation...). Computers would have very different needs and existence, in a very different "world". If you want to avoid those tricky subjects altogether, I think it is more fun to look at the social aspects of AI: do we want AI at all? Why do we want AI? If we already struggle to treat someone of a different culture, ethnic group, socio-economic background or even gender or age as an equal human being, why would we want to create machines that think (and arguably, feel) like us? We'd never accept them as our equal. So exactly how smart and human do we want our machines? Remember: AI is for life, not just for Christmas.
Might not apply but perhaps AI in counter-acting the effects of social engineering in network security. I just picked social engineering because I have a pretty big interest as the human part is the weakest bit of any network.
I'd do that line of questioning, it sounds less messy. @ the claim that intelligence is an example of emergence: tell that to all the substance dualists (i.e. most people everywhere, ever). I actually agree that human intelligence is emergent (i.e. grounded in matter and the combined result of lots of simple neuological reactions) but it's blasphemy for most people. Very few people have a totally unspiritual view of the human mind, annoyingly. The main problem here - and with the "define intelligence" problem earlier - is that we don't yet fully understand the human brain or how it produces consciousness and intelligence. Defining artificial intelligence is going to be a bit tricky when we can't even explain our own. C'mon, neurology...hurry up and fix it all for us...
