I’ve heard this argument many times over, and I tend to agree here. If we think about AI as simply a tool to leverage for our jobs, then we can liken it to the introduction of the calculator. I like this analogy, as it transcends more than just jobs, there is also the impact these tools have on our education systems. Admittedly, it was well before my time, but I am led to believe the introduction of calculators was met with resistance from educators. They were worried that calculators would make us dumber, as we wouldn’t need to think or learn as much (sound familiar?), and students would cheat. Reluctantly, curriculums changed as did examination methods. By the time I reached high school scientific calculators were mandatory, and our math curriculum was based around using the tools available. Recently, I was discussing this exact topic with a university lecturer, who shared a similar outlook. Instead of fighting the tools and technology, she must now adjust her teaching methods to maximise the use of such tools.
My first real ah-ha moment with AI as a tool was along these lines. I needed to interpret a piece of code from a language I was unfamiliar with and converted it to another language I was a little more familiar with (but still not 100% confident). I started to break the code apart and attempt to interpret it, as this was around the time ChatGPT was gaining traction, I figured I’d see if this new tool could help me. Indeed, it broke apart each significant line, explained what was happening, and then succinctly summarised the entire snippet in easily digestible terms. This simple task saved me significant time. Then I started to rewrite the code in the updated language, after I realised I needed to google how to do a task or two, I realised I had a tool that could do this for me. ChatGPT then provided me with the translated code. Ok, so I still had to sanity check the results, but in this case this tool not only saved me hours of work, but my client also benefited as it saved them both time and money.
I’ve spoken to others about this and they too have been able to list off the various AI tools they have used to help in their jobs, to augment their current skills and experience. The feedback I’m hearing is about the positive impact AI has had on their workflow, and how it takes away from the tedium of the basic tasks.
If we take for instance Microsoft’s current general-purpose offerings. They’ve branded their tooling as “Copilot”. First, there was Github copilot, they’ve also got copilot in Power Platform, Microsoft 365, Dynamics 365 and now Windows (Announced at Microsoft Build2023). Surely there is a reason they chose the term “Copilot” rather than simply Pilot or LeadPilot or Captain. They are marketing these simply as tools to help you with your day-to-day tasks. You are still the lead pilot, they are simply there to ensure you are efficient, and to help out when needed.
I solve problems
I’ve had many job titles in the past: Software Developer, Software Engineer, Solution Architect. These days I tend to think of myself as Chief Problem Solver. With this in mind, the main tool at my disposal is software but it is by no means the only tool available. As discussed above, AI is just another tool for me to utilise to solve problems.
Many aspects of my day-to-day job don’t involve writing software. Talking to clients, analysing problems, debugging issues, and documentation. Sure the main output of my job is running software, but it is by no means the main part of it. As it currently stands AI is not capable enough to handle these more complex domains. I’m not saying that it never will be, in fact, it may be likely that in the next 2 - 3 generations of AI, these types of problems will be addressable.
This threat is not new
Ever since I started in the industry, almost 20 years ago, my job has been under threat from outsourcing to cheaper parts of the world. Yet, I’ve still managed to maintain a sustainable career. I know that in Sydney, Australia software development isn’t cheap, some organisations choose to outsource their work to other parts of the world, whilst others choose to keep it here. I’ve heard differing reasons for this: timezone issues, language and/or cultural barriers, quality of work (Although, I believe this is a myth. You don’t always get what you pay for) etc. The reality is that in its current state, AI cannot negate these barriers sufficiently enough to surpass the outsourcing companies and take my job.
I don’t know what I want to do
As discussed previously, at the core of it, I don’t really know what I want to do as my job specifically. However what I am currently focusing on is my purpose, of helping others realise their maximum potential. If I focus on this purpose, then my “job” definition is free to change as it needs to. As discussed in my previous post, 30 years ago, my job didn’t exist, software was an emerging field. Likewise, the jobs of tomorrow are also emerging. Today I may be a Solution Architect, tomorrow? Who knows, and I’m comfortable with this fact. If I focus on my purpose, I know I cannot be replaced by AI anytime soon.
So, am I concerned that AI will take my job? No, not really. I will evolve and grow just like the AI tools will. My challenge is to grow and stay relevant ahead of the curve. This has always been the challenge of my career, as such, I’m looking forward to the challenges ahead, especially with the shiny, fancy new tools available to us.