It is commonly known that AI stands for artificial intelligence. It is best known as the blackbox that organizations like to tell their clients. Startups can seemingly secure funding by simply mentioning the words “artificial intelligence.” But what does it really mean? And how is it used today?
One of my favorite definitions of artificial intelligence comes from the great John McCarthy, who I will discuss in greater detail below, played a huge role in AI as we know it today”
“The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.” — John McCarthy at the Dartmouth Summer Research Project on Artificial Intelligence.
Origins of artificial intelligence
Turing
One of the most famous computer scientists, Alan Turing, defined a Turing test in 1950.
In which he poses the question: “Can machines think?” The test consists of an interviewer in one room and a computer in another. If the interviewer asks questions to the subject and he or she are fooled into believing that the subject in the other room is a human, a machine can pass the Turing test. Turing was a founding father of intelligent machines and showing that they can help out humans immensely.
McCarthy
While the ideas of artificial intelligence existed earlier usually in the form of “intelligent machines,” John McCarthy coined the term Artificial Intelligence in 1956. He was the founder of artificial intelligence labs at both Stanford and MIT in the late 50’s and early 1960’s.
Over the years with increasing computing power and interest in the field, he is one of the main people responsible for artificial intelligence evolving into what it is today.
Applications
A.I. in healthcare
Artificial intelligence is primarily used in three main subfields of healthcare. In order of decreasing popularity in the scientific literature, they are diagnostic imaging, genetic analysis, and electrodiagnosis.
One study from the University of Michigan were able to use new techniques of imaging combined with machine learning approaches to predict brain pathologies with over 92% accuracy.
In another groundbreaking study from Stanford, a convolutional neural network model showed dermatologist-level classifications of skin cancer.
These are all literally life-saving technologies that are sure to increase the average lifespan in years to come.
A.I. in driving
Autonomous cars are becoming more popular within our society each day. The systems promise to remove human error from the equation entirely. In the greatest autonomous vehicle study to date, Lex Fridman and colleagues tested 29 vehicles that are capable of at least partially autonomous cars. There is remarkable technology being created that will be able to monitor driver position and ensuring that both hands are on the wheel.
Such advancements will ensure that humans are not distracted while manning the autonomous car and will cut down on driving while texting.
Additionally, autonomous cars are experts at detecting objects with their onboard camera systems.
As shown in the image above, not only can the algorithm see there are objects present but it can even tell what kind. This type of data during driving is extremely valuable and promises to be part of the next wave of innovation in driving.
Future – Artificial General Intelligence
The future of artificial intelligence promises to be well – interesting. Interesting is the best term that I can think of that can describe it. There is so much innovation occurring with more corporations and politicians getting onboard every day. It will likely save many human lives through earlier cancer detections and safer driving through autonomous cars. However, the continued development of artificial intelligence may one day result in artificial general intelligence.
Artificial general intelligence is loosely defined as A.I. that can take “knowledge” learned from one specific task and apply it to another. Currently, AI is very specialized – meaning that it can be very good at a certain task but can’t exactly generalize. Human manipulation is needed to modify the techniques and input data so that the same type of model can be switched from say reading MRI images to scanning satellite imagery.
Concluding Remarks
Artificial intelligence is fast becoming a cornerstone of our society. It is now even used with stock trading. With the growth of computing power has come some remarkable innovations that will surely improve our quality of life in the future.
