If you're like most people, you've probably heard of artificial intelligence (AI) but you might not be quite sure what it is. My humble opinion is that no one truly knows, not even world-renown experts. What I can say is it is not the image above, although society and media may portray this. I'll attempt to define AI as a field of technology that deals with the creation of intelligent machines. In other words, it's all about making computers smarter to assist us in life.

AI has been around for a long time, but it's only recently that we've started to see its potential applications. There are many different types of AI, but some of the most common ones are machine learning, natural language processing, and computer vision. Machine learning tries to use a computer's hardware to its advantage in solving certain problems, think instead of your brain spending energy to visualize a trend you can create a graph in excel with less effort. Natural language processing (or NLP) is machine learning done on text data, or words, to attempt to mimic how humans assign value to words & the ability to help use a computer to understand language more intimately. Finally, machine learning in computer vision's goal is to use recording devices to compensate for the human eye and perception systems' limitations to help us observe objects more clearly.

To further illustrate what AI is, I'll walk through a bizarre example; how would a handler in a panda sanctuaries use AI to help take care of endangered pandas? Yes... pandas. Those adorable 800 pound bears that are so cute and spark joy in humans around the world. What most people don't know is that pandas are one of the hardest animals to take care of due to a plethora of reasons including but not limited to the amount of food required to sustain their large body mass, habitat destruction, and high panda child mortality rates. Before we get into the technology, most AI problems start with understanding the goal of the project, underlying issue you're trying to solve and what processes exist to help you achieve your goal. In this example with your imaginary tech-savvy panda handlers;

Goal: minimize the effort it takes to keep pandas safe and healthy in sanctuaries

Issue: pandas require a lot of calories to survive combined with a restrictive diet, pandas are known for hurting themselves due to their large size and lack of mobility, and their habitats have been impacted to a point where human intervention is needed

Process: panda handlers wake up early, monitor pandas 24/7 in-person, provide food and oversight while they are active during the day, help the pandas settle down once the sun goes down, and monitor the pandas as they sleep for any health issues (see more)

Now that we've established the goal, issue, and existing process - the fun begins! How could AI possibly assist panda handlers? In AI, we need a few things to be able to make a difference compared to the existing way of handling the problem. In summary - data, data, and more data. Data is a hot-topic word, but in this example we need to be able to measure information so that we can understand whether new ideas actually have a measurable impact on the problem. Once we have data, AI can use its various techniques to analyze the data more efficiently than humans for certain problems and understand if certain processes are better than others.

What data could we use with pandas? One comes to mind very quickly and mentioned earlier - image data. By using the existing infrastructure of cameras ($0 capital required) and having data scientists create an algorithm that can take the 24/7 panda cams to see when panda is in distress could allow pandas to not require 24/7 human supervision and only notify handlers when issues occur. Furthermore, analysis can be run to understand if there's a certain tree that pandas are falling out of more than others and it could be removed.

That solves the pandas falling off trees, but what about the 20-40 lbs of food needed to feed a panda? The idea for that is to use biometric technology (think an Oura ring, but for pandas) to monitor the metabolisms of the pandas. An AI algorithm can be trained to understand how much to feed the panda to ensure we are not overfeeding (minimizing costs) but also understand exactly how much each individual panda needs to eat per day. Imagine being able to understand exactly how many more lbs of bamboo panda A needs versus panda B... assigning quantitative metrics to gluttony, crazy idea huh?

By combining both of these AI techniques into the process, sanctuaries would be able to operate more effectively and panda lives would be positively impacted. Before I receive comments on the unrealistic nature of this example, I recognize the feasibility + scalability of this idea would likely cause it to not be a viable solution. The idea is not to create the best solution, but to provide a framework for how apply AI to any problem. I'd love to hear any perspective on this example or any creative ideas you may have where AI is applicable!