The article starts by blaming AI for the reduced food menu, a speculative claim which the author made no attempt to validate and which is almost certainly incorrect. I stopped reading right there.
In reality, when getting out first to market, it might be difficult for "AI" to decipher if a user added 1 of 5 available sauces to their chicken wings, so to reduce the likelihood of this error, you remove it until the technology is more mature. Speculative sure, but a strong assumption, and I doubt Mashgin would confirm this.
Its definitely wrong - I've used these exact checkout systems at places with way longer menus than any stadium has ever or will ever have. Even if that wasn't the case, it would still be way too speculative of premise to be worth seriously discussing, especially when the Occam's Razor "they reduced the menu size because its easier, they have a captive market, and why try to make good food when you can just charge $20 for a beer" explanation is right there in front of you.
When the menu reduction coincides with the introduction of vision-based checkout, I don't think it's an obvious overreach to link the two together. It may be right, it may be wrong, but this wasn't journalism, just a guy's experience, and the root cause of that decision doesn't change what the article is actually trying to communication.
