physiology websites for medical students. (Though I can’t say I’m all that thrilled about this one).
I know that the biggest change in machine learning is that we now have the ability to train a computer program on the results of our own observation.
From that observation we can learn all kinds of things about ourselves. A computer program can watch us talking in class and learn what words make us the most anxious and then can find ways to calm us down. We’ve seen that at times, typing a single word (like “fuck”) into a Google search can pull up the top 9,000 search results, showing just how many words people use every day that show up in the top searches. From the Google search results, we can learn all kinds of other things about ourselves, such as how much we talk about dieting, how we deal with depression, what drugs people use to self-medicate and what topics we’re interested in, even what year you were born.
We’ve seen how well this works when you train Google’s search engine on your search history (I’m not making this up). See if you can’t find yourself in the top results, once you take your search history out of the way. So how do we use this to our advantage? Can we program machines to see beyond the text? Can we program them to see who we really are, beyond our data? Researching the Science of Machine Learning and Spying The machines can see us; they’re a reflection of our own actions. Of course, they are. We’ve already seen how humans tend to place emotion in context, our inclination to interpret words in different ways to help the situation.
What about emotion? What do we know about emotion that computers might never be able to understand? What if we could figure out a way to program the computers with the same emotional intelligence we humans have, and take their lack of emotion as a strength? The concept of intelligence is a matter of using knowledge to come up with a solution, a better solution. If machines are going to be truly intelligent, they must do the same thing. In this field of study, a great deal of research has been done, focusing on computers. As mentioned, we can train computers to watch our behavior and decide what makes us the most nervous, depressed, or happy. We can also program computers to learn language and be able to automatically create emotions from text. It’s no accident that the last research field in Machine Learning was Language Processing (or Linguistics). This isn’t some odd discovery.
They’ve been studying human language since the 1800’s. Now how does that help us? As much as I hate to be negative, I think it actually is a little frightening to know that these “smart machines” (what should really be called “algorithms”) have not only been built on the principles of human intelligence, but they’ve actually been studied by those in this field of study for decades. I have a lot of faith in the intelligence of humans, that we as humans are driven by our emotions
0 Comments