Yeah so let's refocus on the real perspective here. It doesn't matter at all how good these models are at code or finding vulnerabilities if we destroy our ability to seek and share knowledge.

https://futurism.com/artificial-intelligence/google-ai-overviews-misinformation

Analysis Finds That Google’s AI Overviews Are Providing Misinformation at a Scale Possibly Unprecedented in the History of Human Civilization

A new analysis commissioned by The New York Times suggests that Google's AI Overviews are wrong an astonishing percentage of the time.

Futurism

@mttaggart @lisamelton i find that 100% of the time for subjects I know Google’s AI summaries are flawed (often dramatically so but always in at least some nuance or details)

Which leads me to conclude that it is highly probable that it is equally flawed in topics I’m not an expert.

And google search - even without the ai summaries is deeply flawed today I constantly see search results with summaries of a link that has no correspondence with the actual link (as in the link is something else)

@mttaggart @lisamelton this is especially the case for Google shopping which half the time for the stuff I’m searching shows links to sites that no longer have the item linked in stock, have it for a different price or don’t even show the item in question on the site at all anymore. And Google shopping has seemingly no concept of editions - showing results for say a print on demand version of an old book alongside 150 year old 1st editions of the same work

(it isn’t just Google eBay also)