Even if McDonalds could tell Google not to show the ad to vegetarians, they have no way to verify Google isn't.

At the scale of a McDonalds ad campaign that seems easy to measure: go to the nearest Target parking lot and ask as many people as you can two questions: "are you vegetarian" and "did you see {description of ad here}". Then compare how which percentage of vegetarians saw the ad compared to the percentage of non-vegetarians saw the ad. There are even companies specialized in market research that will do this in a more representative and statistically rigorous way

▲

IAmBroom a day ago | parent | next [-]

Google has no incentive to do so. McDonalds does, but its somewhat costly to scale against ALL the ads they push.

	▲	jmalicki a day ago \| parent [-]
		For pay-per-click ads, if the vegetarians won't click, that's a lost opportunity to show a veggieburger ad Google would have made money from. Google has a lot of incentive to optimize this, and it's a lot of why they (and Meta) spent so much money on AI research that eventually brought us modern AI.

▲

pocksuppet a day ago | parent | prev [-]

infinite things could confound that analysis

▲

derektank a day ago | parent | prev | next [-]

Doesn’t Google earn more from McDonalds if Google provides McDonalds with a higher click through rate? If you’re providing a high click through rate, you’ve probably done the due diligence to screen vegetarians already

▲

mapcars a day ago | parent | prev [-]

Can't they ask claude to create account do some relevant searches and then check if ad is shown?

▲

0c3ca83 a day ago | parent [-]

What makes you think Google has built a profile that shows Claude is a vegetarian?

	▲	mapcars a day ago \| parent [-]
		That's the thing - ask claude to do requests to build such profile