Twelve years ago, when Bart Simpson was writing “I will Google it before asking a question” on the chalkboard, I let Google rule my life (and baby Mark’s) for a day—because apparently Google could solve every problem.

It proved entertaining.

But now it’s 2025, and Google is completely useless. The new top thing is AI, from conversational bots to AI-enhanced search results and AI customer service—hell, I had an AI interview for what is probably an AI job contract this week.

I used to wonder if there’s an adult in the room. Now I wonder if there’s a human in the room.

So, it’s time to let AI rule my life again. Logical move, right? I’m being replaced by it anyway—might as well see how good it is.

AI as my new travel advisor

“Hey ChatGPT, help me travel!”

Back in the day, Feng and I carried paper copies of Lonely Planet to find hotels and figure out how to get from point A to point B. The advice was amazingly accurate—at least in Central and South America.

We took a boat at 9:45 a.m. on a Tuesday morning because Lonely Planet said so—and there was a boat. We waited for the daily bus to nowhere in the middle of the jungle, half reconsidering our life choices, when a muddy chicken bus showed up.

Crazy.

These days, Lonely Planet guides aren’t as helpful. They try to appeal to everyone, I guess, and the “Accommodation” section is almost irrelevant now that everyone books online.

So how good is ChatGPT as a travel advisor?

When ChatGPT is “thinking longer for a better answer,” it’s really just browsing Reddit and the web, aggregating information and presenting it neatly.

At first glance, it looks awesome.

The problem is, there’s a strong “easier said than done” vibe to its advice.

I’ll spare you screenshots, but ChatGPT spits out info without context or warnings—suggesting remote towns, terrible transportation ideas (who wants to sit on a bus for 52 hours straight?) and exhausting itineraries. At one point, I reminded it that a certain night bus isn’t advisable because of frequent robberies… and it went, “riiiiight…”

I wouldn’t trust ChatGPT for travel advice, much like I don’t trust travel bloggers who haven’t actually been there—or were clearly paid to promote a place. It can aggregate data, but you still have to double-check everything.

Also, it produced the most hilariously wrong map I’ve ever seen.

Dear AI, Fix My Writing (But Don’t Rewrite My Soul)

My toxic trait? Always thinking I’m not good enough—and this extends to languages I’m pretty sure I speak, including French at this stage.

I take writing seriously, but I often second-guess myself. Do people actually say this, or is one of my languages interfering again?

As someone who makes sentences for a living, I renew my subscription for Antidote every year. This neat Quebec-made software spots grammar and spelling mistakes in both English and French, highlights inconsistencies, and even fixes typography for people like me who hate inserting non-breaking spaces manually.

Antidote isn’t always right, but as long as you keep your human eye on it, it’s excellent.

Now, Antidote offers AI-based reformulation and retouch options—so I tried them to see if I could sound 100% native.

Spoiler, I didn’t.

Let’s just say AI reformulated my sentences so thoroughly that I barely recognized them. The result wasn’t better—just bland or foreign to me. The human imperfection that makes writing personal was gone.

My Job Interview With a Robot

In September, I applied for a freelance translation job and was invited for an interview—presumably to prove that I can actually speak both French and English.

Until recently, an interview meant an office visit or at least a video call.

Since it’s 2025, I clicked the link and discovered I was being interviewed by AI.

Awkward.

Let’s take a minute of silence for HR professionals who are being replaced by chatbots. Mind you, AI is replacing many professionals these days—will we ever see an AI bot interviewing an AI translator?

I went along for the experience. It was painful. You can’t joke, can’t build rapport, can’t read body language.

At the end, AI asked if I enjoyed the experience. I didn’t lie—why would I? It has no feelings, and I don’t want the contract anyway.

Humans Still (Barely) Necessary

I’m not against AI. It’s a great tool for aggregating data and automating tedious tasks.

But I can’t trust it blindly—and I’m afraid many people do or are encouraged to.

I’ll wait outside with my flawed human brain, trusted sources, and judgment skills.

Let me know when AI develops those—or at least a sense of humour.