The Good, The Bad, & The Ugly with Chat GPT grading
I received notice from SGC that my grades are done. All cards submitted I first ran through Chat GPT to see how the AI grades stack against the real thing. Pictured are 5 “highlights” from a 15 card submission.
All other cards not pictured were exactly as AI graded or within .5 of the AI grade.
The AI grades first given for the pictured cards:
Joe Burrow - 9.0
Dart Green Disco - 9.0
Waldschmidt: 10/9.0
Griffey: 9.0 -9.5
Acuna: 9.0 - 9.5
As you can see, some were better. Some worse and the Acuna waaaaay off. I’m thinking Chat GPT may struggle with the black borders. Gonna crack the Acuna and take the $15 loss on that experiment. Overall though, Chat GPT seems like a good tool, but not a guarantee.