As the computer revolution has progressed over the decades, computers have become easier to use and more reliable. In parallel, people often feel less of a need to use the services of IT professionals. Let’s break down the reasons why customers resist hiring IT workers:
- The primary reason is the cost of hiring tech professionals, despite IT expertise costing about the same a plumber. There’s little hesitation when a plumber needs to be hired, yet customers often turn to laymen on social media for free tech support. Unfortunately, much of the IT advice on social media is amateurish at best, and sometimes exactly the opposite of what a professional would do.
- People feel that hiring an IT worker will take up too much of their time. I’ll revisit this point in a moment.
- Some people (primarily men) are too macho to ask for help and want to tackle their problems themselves.
- The Apple Store provides free support. In fact, this support is quite limited to cookie-cutter tasks focused on Apple’s own OS and apps. The Genius Bar isn’t going to help you with third-party software, IT projects, networking, network storage, and anything beyond the simplest security controls.
- A recent development is that customers will use AI chatbots to help them solve IT problems.
So can AI help with IT? It sure can. I regularly read Google’s AI overview when I have an IT question. But there is a huge difference between a professional and a novice in interpreting the output of an LLM. I use AI to save time. Novices use AI because they frequently don’t know how to approach a problem at all. A novice wouldn’t know if an LLM provides an incorrect or harmful answer. Professionals can put the LLM output into the proper context.
Some of my clients have used an LLM before hiring me. In the most extreme case, my client admitted that he worked many hours with an LLM to address a problem, without success. It took me 30 minutes to solve his problem, without the use of AI.
While I don’t have the details of precisely how my clients used AI to help with their diagnostics, an agent that can view and interact with their screen might have helped them work towards a reasonable solution. I suspect they weren’t using agentic AI. However, I would strongly warn laymen to think twice before putting your Mac under the control of an AI agent. Chatbots currently provide incorrect answers too frequently. Chatbots will inevitably become more intelligent and more accurate over time, but if a chatbot sends you down the wrong path 20% of the time, that’s not a risk worth taking. Also, when you use an AI agent you are taking a privacy risk.
I view AI use in the IT field as analogous to any other profession. Would you let an AI diagnose your medical problem or perform surgery without the oversight of a physician? At this time, agentic AI is best used to automate processes that you already understand how to do. Limiting the scope of what an agent can see and act upon is important. Giving an agent carte blanche access to your Mac is risky at best, and can be outright dangerous.