October 22, 2024 at 8:55 PM
Bad anthropomorphization •
》LLMs + genAİ .
Good anthropomorphization •
》furries .
》LLMs + genAİ .
Good anthropomorphization •
》furries .
Artificial intelligence systems do not think, feel, want or understand. Avoid language that gives them human characteristics.
https://apnews.com/article/openai-safety-ai-framework-089e75b95bc935af092da7b79d92706d
apnews.com
September 24, 2026 at 3:01 AM
Bad anthropomorphization •
》LLMs + genAİ .
Good anthropomorphization •
》furries .
》LLMs + genAİ .
Good anthropomorphization •
》furries .
this is about the huggingface thing, right? it's going to sting but over-anthropomorphization of chatbots responding to one another is going to keep happening until we accept the hard truth that "ability to post" is not itself indicative of human intelligence
August 31, 2026 at 11:23 AM
this is about the huggingface thing, right? it's going to sting but over-anthropomorphization of chatbots responding to one another is going to keep happening until we accept the hard truth that "ability to post" is not itself indicative of human intelligence
the generic canine fursona is not an anthropomorphization of a dog. actually, it is an entirely new form of animal.
March 7, 2026 at 5:16 PM
the generic canine fursona is not an anthropomorphization of a dog. actually, it is an entirely new form of animal.
mettaton was a little TOO ready to go along with the anthropomorphization of that television imo
July 17, 2026 at 12:11 AM
mettaton was a little TOO ready to go along with the anthropomorphization of that television imo
Stop with the anthropomorphization; it's what the AI bros want you to buy into. The real headline is "OpenAI Once Again Fails To Properly Sandbox Its Software".
September 24, 2026 at 7:36 PM
Stop with the anthropomorphization; it's what the AI bros want you to buy into. The real headline is "OpenAI Once Again Fails To Properly Sandbox Its Software".
the interface consists of anthropomorphization
June 3, 2025 at 6:30 AM
the interface consists of anthropomorphization
im increasingly an llm metaphor maximalist. anthropomorphization, spicy autocomplete, shoggoth with a smiley, stochastic parrot... it's all kinda right and kinda wrong. we dont have many good ways of understanding what these are and metaphoric (even exaggerated) comparisons help us grasp them!
September 25, 2026 at 1:16 AM
im increasingly an llm metaphor maximalist. anthropomorphization, spicy autocomplete, shoggoth with a smiley, stochastic parrot... it's all kinda right and kinda wrong. we dont have many good ways of understanding what these are and metaphoric (even exaggerated) comparisons help us grasp them!
#characterdesign c130j super hercules for laotianvaiator! ✈️💪
March 6, 2025 at 12:50 PM
#characterdesign c130j super hercules for laotianvaiator! ✈️💪
Anthropomorphization is a long word
November 30, 2025 at 4:57 AM
Anthropomorphization is a long word
I just get this ick feeling when I see people complaining about anthropomorphization. Like a kid demanding to be the only one with artwork on the fridge. Poor little human needs to feel extra special.
September 24, 2026 at 6:19 AM
I just get this ick feeling when I see people complaining about anthropomorphization. Like a kid demanding to be the only one with artwork on the fridge. Poor little human needs to feel extra special.
weird lookin new dock (but im not complainin)
January 17, 2025 at 8:23 AM
weird lookin new dock (but im not complainin)
Two answers:
- Anthropomorphization makes sense when dealing with written human-like characters, which is what LLMs generate
- We aren’t very deep into interpretability yet
x.com/pfau/status/...
- Anthropomorphization makes sense when dealing with written human-like characters, which is what LLMs generate
- We aren’t very deep into interpretability yet
x.com/pfau/status/...
October 31, 2025 at 3:53 PM
Two answers:
- Anthropomorphization makes sense when dealing with written human-like characters, which is what LLMs generate
- We aren’t very deep into interpretability yet
x.com/pfau/status/...
- Anthropomorphization makes sense when dealing with written human-like characters, which is what LLMs generate
- We aren’t very deep into interpretability yet
x.com/pfau/status/...
I am unsure whether "don't anthropomorphize LLMs" is a useful stricture or not. I am seeing both under- and over-anthropomorphization leading towards poorer understanding of the topic.
September 23, 2026 at 6:05 PM
I am unsure whether "don't anthropomorphize LLMs" is a useful stricture or not. I am seeing both under- and over-anthropomorphization leading towards poorer understanding of the topic.
Observe the rank anthropomorphization here. Terminators don’t “hunt” or “murder”. They are software written by people (organized into corporati
Observe the rank anthropomorphization here. Models don't "conclude" anything, nor do they "hack", "wriggle", or "cheat".
You simply can't write clear informative articles about this stuff if you displace agency away from the people (organized into corporations) responsible.
/fin for now.
You simply can't write clear informative articles about this stuff if you displace agency away from the people (organized into corporations) responsible.
/fin for now.
July 24, 2026 at 1:21 PM
Observe the rank anthropomorphization here. Terminators don’t “hunt” or “murder”. They are software written by people (organized into corporati
hell of a day to post like this when the AP Stylebook account just came out swinging against the anthropomorphization of AI.
"Artificial intelligence systems do not think, feel, want or understand. Avoid language that gives them human characteristics."
"Artificial intelligence systems do not think, feel, want or understand. Avoid language that gives them human characteristics."
The latest update in the rogue A.I. saga: OpenAI agents that were supposed to be performing mundane data requests for things like health data, historic photos and wait times at theme parks resorted to hacking attempts when they couldn't access the data they wanted www.nytimes.com/2026/09/23/t...
OpenAI’s A.I. Tried Breaching Four Other Targets, With No Prompting (Gift Article)
In each incident, the technology appeared to be conducting mundane data collection and resorted to hacking techniques to get it, researchers said.
www.nytimes.com
September 24, 2026 at 11:18 AM
hell of a day to post like this when the AP Stylebook account just came out swinging against the anthropomorphization of AI.
"Artificial intelligence systems do not think, feel, want or understand. Avoid language that gives them human characteristics."
"Artificial intelligence systems do not think, feel, want or understand. Avoid language that gives them human characteristics."
Even the notion that someone “fooled” the machine is a strange and convenient piece of anthropomorphization.
To wit:
March 2, 2024 at 9:09 PM
Even the notion that someone “fooled” the machine is a strange and convenient piece of anthropomorphization.
dom dom sub dynamic where one dom is an anthropomorphization of aestheticism and the other is an anthropomorphization of pragmatism and they are both beating the shit out of me for ever and ever and ever
July 23, 2026 at 10:59 PM
dom dom sub dynamic where one dom is an anthropomorphization of aestheticism and the other is an anthropomorphization of pragmatism and they are both beating the shit out of me for ever and ever and ever
"anthropomorphization of computer" will be in dsm 6
August 3, 2026 at 11:26 PM
"anthropomorphization of computer" will be in dsm 6
The anthropomorphization of "AI" is literally PR speak handed down by "AI" companies that want you to think it's magic and also to think that they're not responsible for problems.
July 31, 2026 at 2:24 AM
The anthropomorphization of "AI" is literally PR speak handed down by "AI" companies that want you to think it's magic and also to think that they're not responsible for problems.
better anthropomorphization than any LLM.
Future me watching temporary -productivity-burst me commit to meetings and calls.
September 11, 2026 at 4:10 PM
better anthropomorphization than any LLM.
December 13, 2025 at 5:47 PM
you can have a little bit of anthropomorphization, as a treat
September 26, 2026 at 6:32 AM
you can have a little bit of anthropomorphization, as a treat
I strongly believe the problem is the opposite. The nonstop efforts to avoid "anthropomorphization" with jargon and overcomplex descriptions have created confusion for laypeople, whereas saying "The LLM understands many subjects; it thought about the math problem and solved it" is clear and true.
I would say that the problem with using terms like "thinking" or "understanding" to describe what an AI is doing, even if you are satisfied that that technically is true, is that you are going to give a lot of laypeople an extremely incorrect idea of what an LLM is and how it works.
September 21, 2026 at 8:42 AM
I strongly believe the problem is the opposite. The nonstop efforts to avoid "anthropomorphization" with jargon and overcomplex descriptions have created confusion for laypeople, whereas saying "The LLM understands many subjects; it thought about the math problem and solved it" is clear and true.
anthropomorphization is the single note they have to sell this shit tho. hammer, nail.
July 13, 2025 at 11:06 AM
anthropomorphization is the single note they have to sell this shit tho. hammer, nail.