WWCode Talks Tech #26: An Overview of Artificial Intelligence

Women Who Code Talks Tech 26     |     Spotify - iTunes - Google - YouTube - Text

Elizabeth Addo, Lead at Women Who Code, shares her talk, “An Overview of Artificial Intelligence.” She shares the history of AI, the different types of AI that exist, and the many education and career opportunities available in this field.

Artificial intelligence is computer science concerned with the automation of intelligent behavior. According to Webster's dictionary, intelligence is defined as the ability to learn, understand or deal with new situations. Further definitions for AI tell us that its algorithms, methods, and techniques enabled by constraints suggest models targeted at thinking perceptions and actions. Artificial intelligence has existed for about 80 years and still counting. Artificial intelligence started between 1940 and 1956, known as the gestation period. That is when McCulloch-Pitts developed the Boolean circuit of the model of the human brain. Alan Turing introduced the Turing Test to test the intelligence of machines. That was followed by the information theory and then a downward meeting. 1956 and then the late 70s is known as the rise of AI. We had the early machine translations and then the early works in the knowledge of representation.

We had working machine translations in the late 60s to the early 70s. A lot of funding for AI research stopped around that time. More work on expert systems was done in the 70s to the mid-80s, known as knowledge-based systems. Systems based on a specific domain of knowledge were introduced. In the mid-80s, ML was introduced, or ML was retained. They started working on more powerful machines and neural networks. They were using natural language processing and adapting new scientific methods for AI. From then to date, we have seen that AI has advanced a lot. We have a lot of data sets to work with and documents, algorithms, and much more. The foundation of AI is built on philosophy. Philosophy, as in, can be used to draw valid conclusions? Questions like how does the mind arise from the physical brain and mathematics? What formula to use, how to derive, how to add, and economic decision making. Questions like what to do, what not to do, and how it will go.

It is also built on neuroscience, the brain's workings, functions, and information processes. It is built on psychology, the study of how humans and animals think, act, and interact for AI learning. We have computer engineering, which also focuses on building the systems and the control. Cybernetics is how the artifact can operate under its control. We have three types of AI, Artificial Narrow Intelligence, Artificial General Intelligence, and then Artificial Super Intelligence. The ANI is the most basic part of the AI system. It involves Siri, Cortana, Google Assistant, and all the basic AI at our disposal. Artificial general intelligence is also focused on a human-level based system where the machines can reason and act on the human level. We classify robotics and machines being built to behave in the human-level knowledge-based way under artificial general intelligence. Artificial superintelligence is the future of AI that we all see when we mention AI.

How is AI going to be super intelligent to humans? It is going to be that AI systems don't require human effort anymore. AI will do everything, but this is yet to come. Currently, research is being centered on ANI and AGI. Once artificial general intelligence succeeds and becomes stable, they will move to artificial superintelligence, according to research. We have the applications of AI. We have computer vision, which deals with the area where the machines are trained to visualize things they see in their environment. They'll be able to learn and use those kinds of data around them. We have robotics. We have deep learning, the machine's ability to also be at the human reasoning level. Machine learning is the process where the machine is trained with the data it receives. It has to study the pattern with the analyzers and all those kinds of things. The machine is being taught, so it's called machine learning.

Natural language processing is where the machine is taught to understand basic and complex languages, both our natural languages and ones imported into the computer. Logic is also a representation of the formalism of all AI applications. We have some areas under the applications where AI is involved. AI is in tech, mathematics, computer science, and electrical engineering. An AI example is Sophia, the first AI humanoid robot. She's able to communicate on a basic level. She does have a smooth level of communication with humans. She also sings. For now, it's just the upper part of her body; her limbs are not developed. She is the current existing humanoid. Miquela is another example. I discovered Miquela on Instagram. She is a virtual robot. They use AI technology, and she's able to change her appearance. She does everything. She does modeling, she does everything that we all do but with this identity as a virtual robot. We have our basic classified AIs: Siri, Cortana, and Google Assistant. They are classified as voice assistants. Further, we also have our fingerprints from our mobile phones, facial recognition, voice translation, and many other AI tech out there.

What makes AI useful? I believe one reason people are inclined to AI lately is because of how efficient it is. It can do the task that it is assigned to do. We might expect an error, but when it is being made, we know where it is coming from. Compared to humans, it is more advanced and does the work efficiently. It has been programmed; it can make analysis more accurate compared to humans. Because of this nature of AI, it is making it more reliable and it is responsible. We tend to rely on AI more than human efforts because it's an already structured program. Some courses in AI are Data Structures and Algorithms, Mathematics, Probabilities, Statistics, Python, Programming, C++, Java, MATLAB, and Discrete Math. There are a lot of courses that one can take to enter the field of AI. The career prospects in AI are AI engineer, machine learning engineer, robotics engineer, data scientist, research scientist, big data software engineer, and many more. AI is an evolving field. It's an interesting field to explore, with many opportunities.