The Fleet Science Center will be open from 12noon to 9 p.m. on Friday and Saturday, December 6 & 7. Free admission is from 5-9 p.m. Friday & Saturday.
Tech It Out: Artificial Intelligence
by Ana Valeria González, PhD. Student in Machine Learning and Natural Language Processing
Often when someone finds out what I do, the reply is something along the lines of are you building the next Skynet? Haven’t you seen Terminator? Or something like, are you automating all of our jobs? The reality is that when people hear I work in the field of Machine Learning (ML) and Natural Language Processing (NLP), they immediately equate that to Artificial Intelligence (AI). These terms are difficult to distinguish and understand because they have been used somewhat interchangeably depending on whether you work in the industry, in research or writing the headline of an article and want to grab people’s attention. Many businesses, for example, will use AI to refer to their machine learning technologies because it might attract more investors or customers. On the other hand, machine learning researchers have often stayed away from using the term AI in a scientific setting because AI, as the public tends to think of it, is not what they do.
What is Artificial Intelligence?
Often the first thought people have about Artificial Intelligence is related to robots that will end up becoming self-aware and, in a Hollywood-esque type of way, take over the world. This is somewhat misguided. AI generally involves the question of whether machines could do the types of tasks that involve human intelligence such as problem-solving, recognition of patterns, sounds, objects, planning, understanding and generating language, among many others. These things can occur without the need for a physical robot. When people think of AI as portrayed in the movies, they often think of a system that has what is called General AI, a system that can perform ALL the tasks involving human intelligence. In reality, systems can perform only one or a few tasks and in a very narrow manner. For example, we may encounter an AI opponent in a chess game who is extremely hard to beat. This means the machine we are facing in the game is really good at planning the next move. However, this does not mean that this machine can also plan and reason for your finances. You would need another machine that is trained for performing this specific task. This is often the reality of AI systems. These systems do not have General AI, they only have a very narrow AI.