The Good News - AI won't/can't take you job (yet)
For anyone who’s played with ChatGPT, two things quickly jump out. The responses to even vague and obtuse questions are eerily human. Except for the shockingly fast output, you’d be hard-pressed not to believe that there is someone on the other side of the screen hand-typing the response.
The second is that it doesn’t take too many probing and specific questions to begin to hit the limits of its capabilities. Context is lost in short order. The “hallucination” issue emerges, where facts are made up and presented confidently. This begins to expose the difference between predicting based on pattern matching and real understanding.
While these gaps and issues may be addressed, off the shelf, the tools today aren’t ready to replace anyone in the org chart just yet.
The Bad News - Someone using AI will almost certainly take your job
The advantages of using tools like generative AI are already paying off. Beyond the anecdotes, early studies show that people incorporating ChatGPT into a written word workflow, when the domain of their work was in the model’s training set, time to complete tasks decreased by 40%, and quality increased by 18%.
Productivity gains like that are already being realized, which means more formidable competition is here. This pressure will come particularly from new entrants without any legacy process or infrastructure to slow them down.
We can look no further than the advent of mobile computing and the example of peer-to-peer payment made in the banking industry.
The Good News - You're not in a zero-sum game
While productivity gains can help with cost and time savings, the real advantages come in more fundamental change. Doing things that no one can or is willing to do today.
A helpful way to think about this is that AI is the college intern. Smart, capable, and most of all, willing and able to follow directions. However, they lack almost any practical experience and judgment.
When given specific and detailed enough instructions, the intern can add a lot of value. Give an intern something to read, even a long and technical industry white paper, they can present a useful summary. Give them a stack of them, and over their summer stint, they can summarize it all and begin to connect the dots between them.
Now, imagine you have not one intern but 1,000. You also can easily provide all of them with detailed instructions and have them efficiently work together. Now, you have a working model of the type of things that the current AI capabilities provide.
Do you have five people in a retail location or branch? What if they each had ten interns to do menial tasks, write personalized and highly relevant emails, chat with every customer, and read every new article or trade publication in your industry? What if they each had 100 interns? 1,000?
What could they do tomorrow that is impractical or impossible today? How much better could your offering or service level be? What offers could you develop or deliver that don’t exist today?
This is the real opportunity and the place that will significantly benefit those who embrace the tools and, more importantly, the vision of the world that could be using what has been made available to all.