I asked ChatGPT to judge my favourite Pokémon cards, and it proved why collectors matter
I asked an AI chatbot to rate my prized Pokémon cards, and in all honesty, it was far too market-focused for my liking.
ChatGPT and AI in general have become ridiculously prevalent - it won’t be long, sadly, until someone asks you to ‘ChatGPT it’ in the way people ask you to ‘Google’ something. But Pokémon is a three-decade-old franchise with plenty of personality, differences in opinions, and a variety in its audience. So, if people are going to keep asking ChatGPT for opinions, can it judge Pokémon cards as well as a collector can? The results - well, it’s a little too happy for my liking.
When it comes to the Pokémon TCG, I’m likely one of the most positive people in the fanbase. People will swarm on Phantasmal Flames for that Charizard chase, yet avoid Perfect Order or Chaos Rising due to it having ‘less popular Pokémon.’ To me, they’re all cards that I want. I am not picky. I’m very much on the ball when it comes to the cards I like, and I’m not sure if Artificial Intelligence can keep up with me. Sorry, OpenAI.
But the proof is in the pudding. And, I’ve decided to send an assortment of my favourite cards over to the ChatGPT servers for its thoughts. This is between a variety of modern Pokémon sets, so there’s no one-expansion-only solution - I’m really putting it to the test. And, despite my sickening positivity with the chaos right now, even I find that ChatGPT is a little too happy for my liking.
For reference, I sent over N’s Zekrom promo from Ascended Heroes (with the Pokémon Center stamp), Gardevoir SIR from Scarlet & Violet base, Pikachu GG30, Squirtle IR from Stellar Crown, an all-time favourite with the Jacinthe SIR from Perfect Order, and well, a bunch of other IRs I’ve scored. Oh, and it also rated the cards based on appeal, art, and market potential. With that out of the way, let’s start with Ascended Heroes’ ETB promo.
ChatGPT gave N’s Zekrom a massive 10/10. “This is my favourite card you've shown,” the robot happily says. Good choice, I think. It then suggests a ‘review’ from Kanto Post’s own words, saying, “If someone asked me what modern Pokémon card art can achieve, this is one of the cards I'd show them.” I don’t know what that’s supposed to mean. There are plenty of cards with varying appeals to different collectors. N’s Zekrom is not cute, nor powerful in terms of design, so I think this statement is a bit hyperbolic, if not overdramatic.
Next is Gardevoir from S&V, which it says is from Paldean Fates. No matter. ChatGPT gave it a 9.5/10, saying, “This card has become iconic for a reason,” ChatGPT told me, but I don’t fully agree with that assessment. I still think Gardevoir is strangely overlooked compared to cards like Bubble Mew or Team Rocket's Mewtwo ex, which seem to dominate conversations far more often.
Those were the two ‘S-Tier pulls’ in ChatGPT’s technological eye. The next one is proudly in the A-Tier - Erika's Tangela from Ascended Heroes. It still got a 9/10, but I have to agree with ChatGPT when it says, “I'd happily take this over many modern gold cards”, as I have the same opinion - but when it’s taking market value into account, I feel like that’s a bad take. Shame on you.
Rather than break down every card individually, here's ChatGPT's final ranking:
- N's Zekrom
- Gardevoir ex SIR
- Pikachu ex (Terastal Festival) TAG 10
- Erika's Tangela
- Squirtle IR
- Mega Charizard X ex
- Pikachu GG30
- Ethan's Magcargo
- Jacinthe SIR
- Piplup IR
The biggest crime wasn't underrating Gardevoir or overrating Charizard; it was putting Jacinthe ninth. I've spent months telling people that Perfect Order is an underrated expansion, largely because of cards like this. ChatGPT saw a pretty Supporter card, but I see one of my favourite artworks in the entire hobby.
I find myself mostly agreeing in terms of an artwork-led focus, but I think the idea of it including market value and ‘investment value’ (ew) misses quite a bit. N’s Zekrom certainly has a chance at making good money in the future, with Ascended Heroes’ popularity. I’d put Squirtle IR much higher up, and the Mega Charizard X ex that is fairly high is the basic ex card - not the SIR from Phantasmal. The results? I’ll give ChatGPT a ‘nice try’ for trying to emulate my opinions, but it’s missing the mark just slightly.
After reading through the rankings, I realised ChatGPT had a problem that many collectors don't. It was trying to find the "best" cards, and I don’t look at cards that way. My favourite cards aren't always the most expensive or rarest. Sometimes they're just the cards that make me happy when I open my binder. That's why I don't think AI will replace Pokémon collectors anytime soon. It can identify a chase card, but it can't tell you why Jacinthe from Perfect Order means more to you than a card worth ten times as much.
So, I wouldn't recommend using ChatGPT to judge your pulls. Instead, stop paying for an AI subscription and start helping independent journalism by signing up for Kanto Post's newsletter. If you'd rather spend that money on Pokémon cards, you can always join our Discord and let real collectors judge your latest hits instead.
