There’s a lot of noise about AI at the moment. Saaspocalypse Now was a whole day built around it. The government has announced that about 8,700 public service roles will go over three years, with AI part of the reasoning. And then there’s AI slop, low quality content made in volume, which has reached us here in NZ too, with fake NZ news pages on Facebook posting AI images alongside real reporting.
I use AI in my own work every day and have even built a web app or two. I’d say I’m still early in my journey, but I’ve got some views (of course). They’ll probably change, which is why I’m calling this my current take.
1. Be curious and start using it. You don’t need to be good at it on day one. Try it on something small and see what happens. If it doesn’t work, try again or try something else. Stay curious.
2. It’s a tool. It needs your thinking and your review. AI will give you an answer to almost anything, and it sounds confident either way. So the thinking is still on you, the human. Ask good questions. Don’t take information at face value. Challenge what looks right and know when to dig deeper or cross reference with another source. What comes out is a draft, and someone has to check it, shape it and own it. If something matters, check it. If it doesn’t sound right, ask why.
3. Build it into how you work and pick the right tool. The efficiency gains are real, but they come from fitting AI into your workflow, not adding it because everyone else has. Work out where your time goes, then choose the tool that suits that job. A tool that suits someone else’s work might not suit yours.
4. Watch out for AI fatigue. It’s easy to fall into a loop where the tool is doing the real thinking and you’re just arguing with it and correcting it. That’s tiring, and it usually means you’ve lost the lead. Know what you’re trying to achieve and what a good outcome looks like before you start, and steer the conversation from there. Step away if you’re going round in circles. Using it all day, every day, can wear you down.
Points 2 and 4 come back to the same thing: judgement. AI works best when you know what “right” looks like, and that comes from doing the work early in your career. It’s the gap I wrote about in Thinking Out Loud. Worth making sure juniors still get the chance to build it.
Photo by Thomas Lefebvre