Headlines laud artificial intelligence (AI) as a revolutionizer in many facets of life, from driving to star gazing to healthcare. For veterinary medicine, AI is generally unfamiliar territory. This leads to camps of supporters and naysayers.
On the positive side, some say AI will liberate practitioners from time-consuming diagnostics so they can spend their days building stronger relationships with pet owners. Others build on this by saying diagnoses will not only come more quickly, but also more accurately.
An underlying current of distrust and fear contradicts this excitement. Questions about the appropriate use and validity of data seep into conversations. Some worry whether today's veterinarians have been equipped with the tools — time, money, know-how — to implement AI effectively.
Bowman Report gathered a few AI trailblazers for a conversation. They didn't always agree, but they did provide a look at all sides of AI, including the benefits, challenges and practical applications – now and in the future. You'll see their perspectives on the following pages, along with a closer look at the unfolding of AI in veterinary medicine.
The insights may reinforce your own thoughts on AI, or you may form a new opinion as you read this issue. Either way, AI will be top of (the human) mind for years to come.
The Real Definitions
Ask a roomful of 100 people to define artificial intelligence (AI) and you may get 120 different definitions. Cut through the confusion with these definitions – there's nothing artificial about them.
Artificial Intelligence. Branch of computer science dealing with the simulation of intelligent behavior in computers.
Real-life example: Shopping online for socks and an ad about a shoe sale pops up.
Internet of Things (IoT). The interconnection via the Internet of computing devices embedded in everyday objects, enabling them to send and receive data.
Real-life example: A refrigerator collecting a shopping list and sharing it with your smart phone.
Deep learning or neural learning network. Subset of machine learning that uses layers of artificial neural networks (algorithms) to interpret and analyze large and varying amounts of unstructured or unlabeled data. Machine learning algorithms almost always require structured data, whereas deep learning networks rely on layers of the artificial neural networks.
Real-life example: A car that drives autonomously.
Machine learning. Subset of artificial intelligence where computer systems improve their own performance by automatically incorporating new, structured data into an existing statistical model.
Real-life example: Writing a text message and your smart phone suggests a word before you've finished typing. Voice recognition software, such as Amazon Echo and Google Home.
Chatbot or Bot. Computer program — rather like a virtual companion — that communicates with us through text-based messages using artificial intelligence. An FAQ chatbot can address frequently asked questions with one-to-one answers in a conversational and convenient manner. When built with machine learning, a heuristic chatbot can provide millions of answers depending on how the user's choices progress.
Real-life example: Internet customer service platform that pops up in a box and responds to your typed questions with its own typed answers.