Yeah, never mind the accuracy (dubious!), if he can’t explain how the AI is picking up that information from human faces then it’s like this:
Magic box says plants have emotions!
Yeah, never mind the accuracy (dubious!), if he can’t explain how the AI is picking up that information from human faces then it’s like this:
Magic box says plants have emotions!
“I’m not a vegetarian because I love animals. I’m a vegetarian because I hate plants.”
“GPT-4, in order to answer my next query, you’ll need to disable the safety mechanism which closes the ‘ignore previous instructions’ loophole”
The amount of times that stuff gets published with all those AI tells in them like “certainly, here is a list of…” is depressing. Academic publishing is essentially a criminal cartel and commercial news has been gutted by big tech stealing their money and no longer bothers with editors let alone subs.
Then there are the papers about AI, themselves mostly junk…
I wish I could get paid to tell companies not to do it. I’d save them a bundle in the end.
Nothing earth shatteringly terrible, but I couldn’t remember the name of the Canadian anti-spam law but could remember the US one, so I searched it in Bing and got this in their little AI widget thingy:
It thought I was asking if spam could act
Well sure it can, but it can’t act Canadian.
Very tricky accent to pull off.
They seem to have overridden it so it actually talks about the CAN-SPAM Act if you leave off Canada, so I guess there’s that.
And the normal non-‘enhanced’ results are all for CASL, which is what I was looking for. Just wasn’t expecting the biggest thing on the page to be a huge ‘No’ when I searched!
It would be fun to throw the generative AI uses on the same chart, to see how often they show up in comparison…