Wondering about what It is like to study artificial intelligence?, Brief answer: Extremely rewarding, yet often frustrating and not always well-defined. It is beyond cool when you’re able to train an algorithm to detect images of cats, identify sentiment in a tweet, or beat some of the strongest chess players in the world (this one is admittedly a bit harder than the previous examples) as these are humanistic in nature and exciting to see a computer do.
However, the field of AI is relatively new (~100 years or so, depending on your definition) with a recent explosion in the last ~10 years or so due to increased computational power. That being said, tremendous progress has been made but there is a ton of work to be done and algorithms and methods are improving every day, especially with the tech giants all pouring billions of dollars into AI research.
On a day-to-day level, Artificial Intelligence or AI can be extremely frustrating as training your models can be time-consuming and difficult to identify reasons as to why your model isn’t working well. Your ability to identify ways to make your model better definitely improves over time, but it can be frustrating (or exciting!) when you first start out. These models also rely on applied math and statistics and having these tools under your belt makes training your models and identifying problems significantly easier. I’d even argue they’re necessary, but one can often get by using libraries such as Keras with very basic linear algebra knowledge.
There’s a lot of hype surrounding Artificial Intelligence and I honestly think it’s totally warranted. The problem is that it can solve are amazing and there’s only more work to be done!
Even when alot of passion is needed to study and understand the concepts of Artificial intelligence. AI is a field which is a combination of mathematics, statistics, computer science, neural science, human psychology, sociology, biology etc.
If someone really develops an interest in the field and able to understand the basic concepts of the underlying technology, from that point onwards there is a lot of interesting things to learn.
AI practically has application in almost all the verticals such as healthcare and medical diagnosis, exploration of natural resources, factories and manufacturing plants, e-commerce applications and social networking sites, banks, retail outlets etc.
AI has the ability to do almost 95% of the tasks that human do at present. With the advancements in the technology and new research happening, AI may be able to overtake human brain processing capability in another 25–30 years.
So, if you are curious about happenings, want to be in the latest of happening technology and has aptitude of the disciplines mentioned above, AI is your cup of tea
My first interest was cybernetics. But, as an autodidact. Later, after a long and deep search with many questions (and finding the answer) and when I felt it necessary to conform enough to get a college pat on the back (as, in theory, academia can be a good thing – they went off the track, starting with Harvard – but, we’ll get back to that some time), I thought that I needed something that weds mathematics and people. What pray tell? John von had an answer. economics. And, I stressed econometrics and mathematical economics.
Aside: Then, I see all of these youngsters trying to run after that set of things; recall, much of the techniques being used related directly to economic modeling. I, on the other hand, figured out the issues there long ago.
And so, with the credentials (graduate work), I went into operations research (OR). Right away, I was in the computing side of things. Of course, I was doing code (and databases), but not from the perspective of CS and its coding. No, people, domains (they are infinite) are where the good stuff can be found. And, OR deals with more than finance and economics. Engineering (and other disciplines) need the methods and mindsets (order important ;>). But too, the Lisp Machine came around. I took naturally to Lisp. Loved the machines. And, that got me more involved with AI. So, being an autodidact, would I not focus on that? We did some marvelous engineering stuff. And, I got to love computational mathematics.
Guess what? My mind kept telling me that we need something called truth engineering. So, that has been my focus the past decade and one-half. Its focus allows me, from a 1G/2G stance, to look at what has gone on and is going on. Sheesh. And, AI being brought to bear in the mess.
ML/DL, with their constraints, trying to be lordly? Much to discuss.
That is why I mention ideas from Carl, as they can fit, not that there are not other ways to look at that this.
One might say that AI is the core issue, meaning that the discipline allows us to look at things correctly, via hermeneutics (this view has been short-changed as people run after numerics). But, computing keeps getting us off course.
Artificial Intelligence is a young field of study started by various disciplines. It is a multidisciplinary field of study. So no unique definition will fit it completely. So depending on the flavour of the approach, AI may look very different. To this you should add your objectives and how deep you want to go in various aspects. Currently, majority of the practitioners are application developers. They use AI and not develop its algorithms. Hardcore IA develops and improve fundamental algorithms.