AI Told Me to Wait. It Almost Cost Me $2,500.

A real story about a 20,000-point-a-night Hyatt deal in London, a Chase Sapphire Reserve transfer, and the moment I almost let AI talk me out of trusting my own judgment.

I found the hotel.

I knew I'd found it.

It should have been a thirty-second decision.

Instead, it almost became one of the most expensive mistakes I've ever made planning a trip.

I still almost lost it because I let AI talk me out of trusting myself.

That's the whole story, honestly. But let me back up.

Quick take: If you find an exceptional, cancelable points redemption — like a Hyatt hotel at 20,000 World of Hyatt points a night when the same stay runs $2,000–$2,500 in cash — book it immediately. Don't ask AI whether to wait. Award availability can disappear (and sometimes come back), but you shouldn't count on the second chance.

I'm planning a graduation trip to Europe for my son, and if you know me, you know this isn't a casual Pinterest-board situation. I'd been mapping flights and hotels for months. Finding a room that can actually fit three people — two adults and a teenager — turns out to be a lot harder in Europe than it is in the U.S., so when I found a Hyatt in London that worked, I knew it was worth paying attention to. I wanted this trip to happen the way I'd pictured it, without wrecking the emergency savings I've spent years building.

So when it came time to lock in a hotel in London, I did what I do now for basically everything: I opened AI to help me think it through.

I want to be clear about something up front, because it matters for the rest of this story — AI has genuinely been useful for this trip. Comparing neighborhoods, sanity-checking timelines, keeping my sixteen browser tabs from turning into forty. It's a good assistant. That's how I use it across travel, money, career moves, home projects, pretty much everything — a place to think out loud, compare options, get organized. It is not, as I learned, a good decision-maker for the moments that actually count.

Half the reason I opened the Chase Sapphire Reserve in the first place was so I could actually use those points for Virgin Atlantic flights and World of Hyatt hotels instead of just staring at them in an app doing nothing. The Sapphire Reserve still transfers to World of Hyatt at a 1:1 ratio, and this particular hotel normally ran $600-plus a night in cash — so having points already sitting there meant I had a way to make it work without draining my checking account.

For three nights, the comparable cash options in the same area were landing somewhere around $2,000 to $2,500. London does not play — it's an expensive city, full stop, and family-friendly rooms like the one I'd found don't sit around waiting for you.

The award price for this particular hotel was 20,000 Hyatt points a night. Three nights, 60,000 points. Compare that to $2,000-plus in cash, and you don't need a finance degree to see what a ridiculous deal that was. This is exactly why I'd gotten into the points game in the first place — it gave me options a straight cash budget never would have.

I knew what I was looking at. I knew it was good. Instead of just booking it, I asked AI what it thought about timing.

The advice came back calm and reasonable: no need to rush, the trip's still months out, award pricing and inventory can shift, keep watching it.

On paper, that's not bad advice. It's the kind of thing a smart, cautious friend might say. The problem is I already knew better. Some part of me — the part that's watched enough of these deals over the years — knew this was a "book it now, ask questions later" situation. Just book it, I thought. AI didn't talk me into anything new. It just gave me permission to ignore what I already knew.

So I waited.

The next day, the hotel was gone. Not "prices went up" gone — gone from award availability entirely.

I got on Hyatt chat. "I'm sorry, that rate is no longer available for those dates." I called. The phone rep told me the same thing — sold out for those dates. If I'd taken either answer and walked away, that would have been it. The redemption would have just been gone.

What did I just do? I had it. I actually had it.

And I want to be honest about where my head went, because this is the part that isn't really about points at all. I'll be honest - I was almost in tears thinking I had just turned a 60,000-point hotel stay into a $2,500 cash mistake. This trip means something to me. I'm trying to give my son an experience I never had at his age, and I'm trying to do it without gutting the emergency savings I've worked hard to build. For a few hours, it felt like I'd fumbled both at once over something that should have taken me thirty seconds to book.

I kept checking anyway, because what else was I going to do — give up? Later that same day, the award availability came back.

This time I didn't ask anybody's opinion. I transferred my points straight from Chase to World of Hyatt and booked it on the spot, before I could overthink it a second time.

It worked out. But it worked out because I got a second chance, not because I made the right call the first time.

What AI Got Right

To be fair to the tool — it wasn't wrong that award pricing and inventory can change over time. That's true. It's generally decent advice in the abstract. If I'd been asking about a flight six months out with plenty of flexibility, "keep watching it" might've been exactly right.

What AI Got Wrong

Here's the thing about AI — it's making its best guess based on millions of data points about how these situations usually play out. That's genuinely useful, most of the time. What it doesn't have is my specific situation. It didn't know that 20,000 points a night against a $2,000-plus cash rate is the kind of number I don't see twice. It didn't know this was a graduation trip for my son, or that I've been tracking Hyatt London pricing long enough to recognize a genuine outlier when I see one. I was the only one in that conversation with the full picture. AI gave me a reasonable answer for an average situation — mine just wasn't average.

I was mad at AI.

Then I realized I was really mad at myself.

AI didn't do anything wrong, exactly — it just did what it's built to do. It gave me the most reasonable-sounding answer it could, based on everything except the one thing that actually mattered: I already knew what I wanted to do. The real mistake wasn't the advice. It was that I had a strong gut instinct, and I let a calm, well-worded response talk me out of it.

What I'd Do Differently

What I actually took away from this has less to do with points and more to do with trusting myself. When I already know the answer — about a hotel, a job decision, a home repair, whatever it is — I don't need AI to sign off on it before I act.

I can research with AI all day. I can compare, plan, and organize with it. AI is an incredible thinking partner. It's just not supposed to replace my own judgment in the moments I already know the answer.

Practical Takeaways

• If you find something exceptional and it's cancelable — a hotel rate, a flight, any kind of deal — book it first and keep comparing later. You can always change your mind. You can't always get the same rate back.
• Use AI to research, compare, and organize. Don't use it to talk yourself out of something you already know to be true.
• When you already know the answer, treat AI's advice as one more data point — not the deciding vote.
• "Sold out" or "unavailable" isn't always the final word, but don't count on it reversing. Act on what you already know first.
• A good strategy — points, budgeting, planning, whatever the tool — is valuable because it gives you options you wouldn't otherwise have. Use it. Just don't let it replace your own instincts.

Quick Questions People Ask Me About This

Should I book a great points redemption right away, or wait for a better deal? If it's an exceptional rate and it's cancelable, book it right away. You can keep watching for something better and change your mind later. What you can't always do is get the same rate back once it's gone.

What's the Chase Sapphire Reserve to World of Hyatt transfer ratio? It's 1:1. Every point transfers straight across from Chase to World of Hyatt, which is part of why the Sapphire Reserve pairs so well with Hyatt stays. (Transfer ratios can vary by card, so it's worth double-checking your specific card's terms.)

Does Hyatt award availability actually disappear and come back? In my experience, yes. My London hotel went from bookable to completely unavailable overnight, then came back later that same day. I don't know exactly why it happens on Hyatt's end, but I've learned not to assume "gone" means "gone for good" — or to count on it coming back, either.

One Last Thing

This won't be the last time I get advice from AI and have to decide for myself what to do with it.

That's what Tish Talks AI is.

Real stories.

Real experiments.

Real lessons.

Honest wins.

Honest mistakes.

I use AI to think better — not to think for me.

If that sounds like your kind of corner of the internet, I'd love to have you along.