ChatGPT Won't Make Us Redundant. Just Stronger

I gave a presentation to the firm a few months ago on the practical applications of dialogue-based AI engines such as ChatGPT. One of the main points I wanted to communicate was that ChatGPT was simply the evolution of the programming languages that preceded it - not unlike how modern JavaScript is an evolution of Assembly. Our interfacing with the Machine finally made the jump from "Machine Language" to "Human Language".

Fred Brooks said that computer programming is the closest we'll get to the magic of myth and legend. Now, instead of writing our magic spells in the arcane language of Computer Programming, we can use plain English - a paradigm shift that will trigger profound changes in our lives. But what does that mean for us?

Information is Cheap. Wisdom is Expensive

The internet democratized information, making access to it redundant. We no longer need to consult with a professional to understand what a Relational Database is - but we do need one to understand whether it's the right technology for our target architecture.

Without understanding a particular subject matter, a layman would not be able to coax the right response from an AI, and more importantly, won't be able to verify the accuracy of the AI's response.

But let's say a perfect AI does exist, one that can magically provide the right answer every time... what then?

The Bar is Higher

Every new technology enables us to do bigger and better things. Technology that seemed to fit squarely in the realm of Science Fiction - such as the touch-screen enabled pocket computers we call smartphones - is now a commodity.

Once AI reaches a level of accuracy and reliability that we consider acceptable - and start using AI to automate most of the tasks we do now - we'll just invent more complicated tasks that the AI can't accomplish, and the race continues - in the same way that the Higher Level Programming Languages automated most of the Low Level Languages' tasks and allowed us to create the wonderfully complex web apps we see today.

There is one more thing an AI can't replace...

Accountability is a Human Thing

At the end of the day, we still want human accountability. If we received a bad advice from an AI - we can only blame ourselves, as the AI is simply following its programming.

Now that we've established that we are unlikely to let an AI take complete control of our lives - let's talk about the next step in our relationship with AI.

Learning "AI Speak" is Becoming Important

Just as any of us can now navigate a smartphone interface with ease - a technology that was not widely available 20 years ago - learning how to speak with an AI is going to become a valuable skill.

Automating and optimizing day-to-day tasks will become the norm once we become comfortable conversing with an AI and understanding the nuances required to produce an optimal result. (In fact, many of us have already started using AI in this fashion - such as setting up alarms in Siri or Alexa!)

The future holds many exciting possibilities for us to use AI - and we'd do well to stay ahead of the curve and learn this technology now.

I am a director with Sionic LLC, specializing in AML technology. Let's have a chat on modernizing your AML monitoring environment! Email me at [email protected].