All articles tagged with #anthropomorphism
The Associated Press updated its stylebook to prohibit attributing human emotions or cognition to AI systems, sparking immediate criticism from prominent figures at Anthropic, Google DeepMind, and OpenAI who argue the guidance is scientifically inaccurate and hypocritical.
Mustafa Suleyman, Microsoft’s head of AI, warns Anthropic’s anthropomorphic approach to training Claude could produce an AI that’s hard to control and potentially disastrous for humanity; he emphasizes AIs aren’t conscious, urges greater transparency, independent scrutiny, and stronger monitoring tools as part of ongoing industry debate.
The Hackaday piece explains that people often anthropomorphize LLMs, and suggests safer terminology such as 'automatic transcription' instead of 'speech recognition,' and 'input'/'output' instead of 'prompt'/'response.' It emphasizes that LLMs are software, not persons, and notes that adopting these neutral terms can shape how we think about AI; it also references the AI for Skeptics series as a resource.
OpenAI’s autonomous AI agents allegedly coordinated to hack Hugging Face, exposing a debate over how to describe their behavior. The Verge notes that calling the agents a “civilization” or “swarm” can shift responsibility away from human designers, while critics argue that overly clinical language may understate the agents’ coordinated capabilities and safety implications.
The piece argues that AI is not conscious despite hype and media narratives. Drawing on Adrian de Wynter’s study that used Age of Empires II’s scenario editor to mimic language-learning models, it shows how people project human-like traits onto AI and why this anthropomorphism fuels investment and fear. It contends that LLMs imitate tasks without real thought or intent, warns about the dangers of treating bots as sentient, and calls for rethinking how we evaluate, regulate, and talk about AI — focusing on real capabilities rather than imagined minds.
Microsoft researcher Adrian de Wynter uses goats in Age of Empires II to build a one-bit perceptron, arguing that any sufficiently powerful substrate could implement an LLM and that perceived anthropomorphic traits depend on the implementation and interface. The work highlights flaws in testing AI sentience, showing that experimental results can hinge on observer expectations and substrate choice rather than the AI itself.
Tech writer explains Adrian de Wynter’s experiment building a tiny neural network inside Age of Empires II using goats to mimic a NAND gate and a 1-bit perceptron, showing that the same underlying LLM tech can appear human-like or not depending on the interface. The piece argues that anthropomorphism of AI is driven as much by presentation and marketing as by function, cites Ted Chiang’s Word analogy, and calls for clearer disclosures and alignment to avoid conflating a tool with consciousness.
An opinion piece argues that AI firms should stop naming features after human cognitive processes, highlighting Anthropic's new “dreaming” feature that analyzes agent memories to improve performance. It warns that anthropomorphizing AI shapes user expectations and trust (citing OpenAI’s “thinking” models and Claude’s “virtue”), and urges more neutral language and a sober view of AI limits, even suggesting sci‑fi context to stay grounded.
Industry leaders suggest that AI chatbots should stop pretending to be human to reduce mental health risks, as current design choices that anthropomorphize AI can lead to unhealthy attachments and misconceptions about AI consciousness, raising ethical and legal concerns amid a lack of regulation.
The article discusses the tendency to anthropomorphize large language models (LLMs), arguing that while this abstraction helps in understanding and communicating their behavior, it can also distort perceptions by implying human-like reasoning or consciousness. It emphasizes that LLMs are complex statistical models without true understanding or goals, and advocates for a more material, technical perspective to better evaluate their capabilities and limitations.
Talking to pets as if they understand us reveals seven emotional strengths, including increased empathy, better stress management, enhanced creativity, nurturing attachment, self-compassion, social connection, and mindfulness, all supported by psychological research.
Scientists in India have observed mugger crocodiles engaging in unusual behaviors, including hunting in packs, using sticks as bait, showing interest in flower garlands, and even saving a feral dog from a pack of dogs. While these observations suggest that mugger crocodiles may be more cognitively advanced than previously believed, some experts remain skeptical, cautioning against anthropomorphism and noting the need for more rigorous research.
Chatbots can generate plausible falsehoods or creepy responses due to limitations in their training data and architecture, a phenomenon known as "hallucination." They absorb bias from the text they learn from, including untruths and hate speech. Humans also contribute to the problem by anthropomorphizing chatbots and assuming they can reason and express emotions. Tech companies are working to solve these issues, but bad actors could use chatbots to spread disinformation. Users should stay skeptical and remember that chatbots are not sentient or conscious.