Talk to your non-AI colleagues too

I interact with LLMs (or so-called AI agents) on a daily basis. As part of my work, I developed several agents that help me proof-check my logic or calculations, and provide quick access to a “rubber duck” whom I can bother at any time.

And that’s been working out quite well. Frankly, I am very considerate of the tasks I delegate or in what way I interact with the agents, leaving the meaningful, skill-building stuff to myself - and the mundane, repetitive, boilerplate tasks to them.

As a young professional, especially in these times, it’s important for me to build my skills as an expert without completely counting on the LLMs. Else how can I truly learn & rely on my experience later on?

However, I was faced with a task at work that left me a bit confused. Pre-LLMs, I would’ve searched the web, or addressed a colleague for clarification once I couldn’t find anything meaningful. But to cut down time & not bother my colleague first, who is probably x10 busier than me, I turned to the agent for help. As I was explaining the situation to it, I realized I was straying away further from the truth. I needed a person to explain me what was going on.

The LLM actually did an okay job as I was somehow still piecing the puzzle together myself, but the final clarification from my colleague cemented that knowledge inside my head. Now I’m much more confident of that concept and can navigate it with more certainty… but this got me thinking.

Imagine, this is happening now - a couple of years after the rise of LLMs. How is this going to look like in 5 years? Actually, forget 5 - how is this going to look next year? Sure, I’m more engaged in the industry than some and interact with AI agents on a daily basis, but people in other areas are picking up pace too. We’re noticing people turning to agents first for help, instead of calling for an expert. I’m not saying that’s necessarily bad, in some situations this might be even acceptable (Did you even Google it, bro?).

However, I wouldn’t want us to lose the human touch. Neither outside of work, nor at work. It should be acceptable to ask a “stupid question”. And asking your human colleague a question, should be acceptable, too.

Thankfully, at work I was always told that there are no stupid questions - all questions are welcome; which is really motivating to keep learning.

So, one thing I realized: talk to your non-AI colleagues too. This sounds surreal and perhaps too early for such a reminder, but it’s good to keep it in mind as we advance further and work as we all know takes a different shape. I’m curious of what follows next.