I was curious so I removed all the text of this image and asked ChatGPT 4o to tell me as much about the image as it could.
The image depicts a man giving mouth-to-mouth resuscitation to a small animal, likely a dog. This is a procedure used to provide emergency breathing support to animals that are not breathing on their own. The man is holding the dog’s mouth open and blowing air into its lungs. Another person is supporting the dog’s body. This technique is part of animal first aid and can be critical in life-saving situations.
“Dog” is a pretty decent guess, without the context I would have assumed it was an otter
Even with the context I thought it was an otter
Fascinating. The model identifies the Harmonica by the hand position.
I guess a catmonica?
What now?
fuck it, cat plainrock124
i thought it was a image edit 💀
I saw an interesting video about image classifiers yesterday.
Harmonicat
Confidence: 58.74%
But the danger daggers!
It says right there that the confidence is way below 50%.
Translation: “I have no idea what this is. The closest thing I know is a harmonica, but I’m pretty sure this is not a harmonica, tho”.
If only the AI shit being crammed down our throats included confidence scores
The AI includes it, the humans choose not to share it.
Yeah, these tools would be a lot better if they could say “I don’t know” more.
Humans too
To be fair, I also wouldn’t be able to tell what it is without the text
Guessing you also know that isn’t a harmonica though?
I’m about 60% sure it isn’t
The hair seems wrong for a cat. Looks more like a sloth, honestly.
My mind went to otter or beaver so it’s definitely not that easy even for us organics.
Hits from the bong
Am I floating yet
Lay the kitty down just for a second
Don’t panic now, it’s the right method
Inhale, exhale