AI is here to stay and it is being received with mixed reactions.

It seems like the young are more “open” to AI, no pun intended, and the tech savvy are finding it very useful. Others do not have a clue what it is and how it works but it is affecting our lives every day and has been for many years. Some just didn’t realize it.

Here are just a few anecdotes that will give you some food for thought.

A Bit Of History

I worked on a software application demo in the early 1980’s that we referred to as “AI”.  That was the first time I had ever heard the term. It was on an early model PC that probably had 32mb of RAM, a DOS operating system that was stored on a 8″ floppy disk and there was no such thing as a mouse. Seventy doctors were given different sets of symptoms from a slew of real patients and asked what they thought patients’ medical conditions were. Individually, they were only correct about 27% of the time.  By entering all of their responses into the computer, and then entering the same symptoms that the doctors were given, the computer was correct over 85% of the time.  That was just a small sample of data collected from 70 doctors.  Today, a doctor can enter symptoms into a computer and diagnose a patient’s condition with amazing accuracy because the computer has stored many thousands of symptoms and end results from many thousands of actual patients. That is pretty much how WebMD works, so we’ve all used AI at one time or other without even realizing it. A rose by any other name ….

Here is one we all can relate to. Everyone has run over two wires at an intersection at some time or other. That is the city traffic department counting cars so they can tweak the traffic lights based on traffic flow at different times of day. That is a form of AI that has been around for many years. They feed that data into a computer and it sets the lights to change accordingly. Rush hour into the city, rush hour out of the city, advanced greens, bi-directional lane arrows being a red-x or green-arrow, etc etc.  Now with live cameras and AI combined, they can do all that stuff real time, without human intervention. So, for example, if a sporting event or concert is causing heavy traffic in a certain direction at an unusual time, the lights can change accordingly, and automatically, to reduce congestion.

Open To Abuse

Today, AI has advanced to the point where students can browse to ChatGPT and enter such things as “write me a 300 word essay on how the knee joint works” and it will spit out a perfectly credible essay in about 20 seconds. Teachers and professors are now forced to arm themselves with software that can read a student’s essay and detect whether it was “plagiarized” using AI or not. Several students handing in almost identical essays no longer means they obtained a copy of the paper from a student who got an “A” the previous year. It means they all entered similar keywords into ChatGPT. So it’s become a game. The student must alter the AI generated essay, with the least amount of effort possible, in such a way to avoid detection and the instructor must try to figure out if they are giving passing grades and fraudulent credentials to students who know very little about the subject at hand.

Who hasn’t seen a video where it looks like a celerity is speaking but what you are hearing is totally absurd? Sometimes it’s a funny joke but it could easily be used to spread believable fake news. Now they have artificially animated news anchors, that look very real and trustworthy, reading the news, but they are just very life like animations that could be used to cut a human salary.

A Useful Tool

As an IT support person, part of my profession involves writing and/or debugging code. Sometimes when new levels of software are released, existing code becomes deprecated and no longer works. If I can determine where the error is, I can copy and paste several lines of code into ChatGPT and ask it to fix it so it works on the latest version of the software. ChatGPT will rewrite corrected code and tell me what it fixed and why and display a “copy” button so I can copy and paste the new code into my program.  I have no doubt I could have fixed the code myself but rather than possibly spending hours doing it, my program maintenance time gets reduced to about 30 seconds, allowing me to move on to the next task on my plate. ChatGPT amasses huge volumes of information and learns from people just like me. Much like the doctor scenario above, it learns on the fly. If I report the fix didn’t work, it spits out a fix to its first fix and we keep working on it until I say “thank you that worked”. Meanwhile, it has stored our entire interaction so it can reference it when the next guy comes along with a similar problem or if I come back in the future to resume the thread. It even says, “You’re welcome. Please feel free to come back when you run into another coding problem”.  I find it weird that I actually use good manners when I interact with ChatGPT, and it uses good manners back. It probably knows it must be interacting with a Canadian. ;-)

A Scary Tool

Last night on television, I watched a segment on how the military is implementing AI on air, land and sea drones. Some of these drones were being used for reconnaissance and such but some were armed and capable of unleashing an arsenal of deadly weaponry. The officers that were interviewed all said “a human must intervene and make the ultimate decision on whether or not to kill”. Somehow I do not feel comforted by those statements, especially after what happened yesterday with Israel blowing up an encampment of Palestinian civilians, admittedly by mistake.  Who or “what” made that mistake? Near the end of the segment, a technology professor from the US said, “I tell all my students, when somebody comes up with a new AI application, there is always somebody else thinking, ‘How can I make a bomb out of this?'”

Synopsis

So yes, AI is here to stay and whether it’s a good thing or a bad thing depends on who is using it and how. Even in military applications, whether it’s good or bad depends on which side you are on.

and so on …

sheff