A lot of you may be wondering what Artificial Intelligence is, I asked myself that question too before watching this expert session video. As Jim Sterne said before in his video, it is NOT science fiction and it is not what you see on TV, as marketers we need to realize that robots and the cliche artificial intelligence that is on TV is so well advanced and well into the future that we don’t need to worry about it at all, and it is highly unlikely too. Artificial Intelligence does covers a variety of subjects and I am here to write about what they are in this blog.
Artificial Intelligence is natural language processing, so being able to hear the voice and turn the voice into meaning using text or voice. Computer vision so you can look at a picture or a video and understanding what is going on in that picture. The most important thing for Artificial Intelligence that marketers can provide is machine learning. Machine learning is a big thing and this is why. We have been studying technology for 50 years but just recently we have enough data and knowledge on this topic and we’re going to use this date to the best of knowledge we can and learn. A lot of it will be testing so trial and error runs will take place so we know which one works best. Machine learning goes hand in hand with artificial intelligence because it has always been an open source, so we can share our ideas and put them out there, kind of like forums. Machine learning is different though, it infers rules from the data in order to create a model then when it does get data it changes it’s mind when new information comes in. Machine learning doesn’t require humans unlike other programs. The machine learning method requires a lot of information though. You need to feed it a lot of data. Just for looking at a letter or a simple picture that you think isn’t that major you do need a lot of data.
In the video he goes over three types of data, there is supervised, non
supervised, and reinforcement. These are all different types of data that Machine Learning has to process. Supervised is knowing the right answer and you will need a lot of examples and labeled data. Non supervised is not knowing the answer the machine has to find patterns unlabeled data, and reinforcement is saying that there is no right answer, some answers will always be better than the next one. In this Expert session there was a lot of valuable points and it showed you pictures and graphs of each type of data and went in depth with what had to be said. This was just an introduction into artificial intelligence, just think about everything all together for this as a whole and we could have so many pages alone just on the key terms. I hope you found this blog post a little informational on this expert session!