If LLMs Have Human-Like Attributes, Then So Does Age of Empires II

2026/05/29 by Adrian de Wynter · 79 voices
#cs.CL #cs.AI #cs.CY

paper · pdf

Abstract

Much research has been carried out on large language models (LLMs) and LLM-powered agentic workflows. However, many works within the field state emergence of, ascribe to, or assume, generalised anthropomorphic attributes to them (e.g., morality or understanding of natural language). Our goal is not to argue in favour or against the existence of these attributes, but to point out that these conclusions could be incorrect. For this we build and train a simple neural network on the videogame Age of Empires II, and note that any entity in a sufficiently-powerful substrate, such as LEGO or the Greater Boston Area, could also present such attributes. Hence, the purported anthropomorphic attributes of LLMs are empirically non-unique: although some properties (e.g., responses to prompts) could remain invariant, others, such as the interpretation of their perceived behaviour, might change with the substrate. Thus, any empirically-grounded discussion on these attributes requires explicit measurement criteria; otherwise the interpretation is left to the representation. We then show that assuming that these attributes exist or not in a system, independent of the substrate and in a generalised way, leads to either circular or uninformative conclusions. This is regardless of the experimenter's viewpoint on the subject, or whether the outcome shows existence or non-existence. Finally we propose a 'null' assumption, where one assumes LLM non-uniqueness instead of assuming anthropomorphic attributes to set up an experiment, along with examples of it. We also discuss potential objections to our work, briefly survey the field, and prove that Age of Empires II is functionally- and Turing-complete.

Citations

• Large Language Models and Emergence: A Complex Systems Perspective
• Why human-AI relationships need socioaffective alignment
• Tell me about yourself: LLMs are aware of their learned behaviors
• Conversational AI Powered by Large Language Models Amplifies False Memories in Witness Interviews
• The Extended Mind
• Does a rock implement every finite-state automaton?
• The perceptron: A probabilistic model for information storage and organization in the brain.

Discussions

• This paper doesn't use the term "intentional stance," but it does really expose that imputation of consciousness is an act by the observer that's usually not based on dispassionate analysis of the obs [bsky, 439 points, 13 comments]
• the GOAT [lemmy, 365 points, 19 comments]
• There's something incredible about an academic paper using the Turing-completeness of goats in Age of Empire 2 to prove that attributing humanity to LLM's requires attributing equal humanity to hordes [bsky, 237 points, 5 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II [hn, 118 points, 121 comments]
• Extremely fun paper demonstrating that if you think your Claude subscription is conscious, you should think Age of Empires II is also conscious arxiv.org/abs/2605.31514 [bsky, 80 points, 3 comments]
• Everyone should read this, just to see how they implemented a perceptron (a very basic neural network) in Age of Empires II arxiv.org/abs/2605.31514 The artificial neurons pass activation energy in th [bsky, 50 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II [lobsters, 35 points, 26 comments]
• Reminds me of this recent paper - “If LLMs Have Human-Like Attributes, Then So Does Age of Empires Il” arxiv.org/abs/2605.31514 [bsky, 24 points, 1 comments]
• Dear fellow Millenials: "If LLMs Have Human-Like Attributes, Then So Does Age of Empires II" arxiv.org/abs/2605.31514 [bsky, 22 points, 0 comments]
• Damn good paper [bsky, 18 points, 1 comments]
• Mad lad trains neural network in Age of Empires II arxiv.org/pdf/2605.31514 @surcomplicated.bsky.social [bsky, 13 points, 5 comments]
• arxiv.org/abs/2605.31514 [bsky, 10 points, 0 comments]
• Fantastic Arxiv find arxiv.org/abs/2605.31514 [bsky, 10 points, 1 comments]
• #MathOfMario arxiv.org/abs/2605.31514 [bsky, 10 points, 1 comments]
• De-anthropomorphizing LLMs via exotic implementations: If LLMs Have Human-Like Attributes, Then So Does Age of Empires II arxiv.org/abs/2605.31514 LLM on a 1979 PDP-11 www.youtube.com/watch?v=OUE3... [bsky, 10 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II [hn, 7 points, 0 comments]
• new age of empires ii hot take just dropped [bsky, 7 points, 1 comments]
• 3/3 arxiv.org/abs/2605.31514 [bsky, 6 points, 0 comments]
• Falls ihr was cooles heute lesen wollt: Hier das paper zum Thema „IF LLMS HAVE HUMAN-LIKE ATTRIBUTES, THEN SO DOES Age of Empires II“ arxiv.org/pdf/2605.31514 [bsky, 6 points, 0 comments]
• How to get me with a title “For this we build and train a simple neural network on the videogame Age of Empires II, and note that any entity in a sufficiently-powerful substrate, such as LEGO or the G [bsky, 5 points, 1 comments]
• Lmaooooo arxiv.org/abs/2605.31514 [bsky, 5 points, 1 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II [hn, 5 points, 0 comments]
• If #LLMs are conscious, so is Age of Empires II. Or: sentience from (virtual) goats. Original paper by Adrian de Wynter: https://arxiv.org/abs/2605.31514 Summary by @404mediaco here: https://www.404me [bsky, 4 points, 1 comments]
• arxiv.org/abs/2605.31514 [bsky, 4 points, 0 comments]
• I run this through Gemini and it seems to be as bad as I thought based on the abstract, so I'm not planning to waste more time on it: "You aren't misunderstanding the paper; the paper is misunderstand [bsky, 4 points, 1 comments]
• yeah a perceptron isn’t big enough to draw a comparison to LLMs imo arxiv.org/abs/2605.31514 [bsky, 4 points, 1 comments]
• Next time someone tries to tell you an LLM is conscious, send them this article by a principal scientist at Microsoft: If LLMS Have Human-Like Attributes, Then So Does Age of Empires II arxiv.org/pdf/ [bsky, 3 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II [hn, 3 points, 0 comments]
• FINALLY finished reading this one. It is so, so, so, so GOOD! arxiv.org/pdf/2605.31514 And funnily enough, I had it open in six different tabs, so I get to close six tabs! [bsky, 3 points, 1 comments]
• Here’s the paper! arxiv.org/pdf/2605.31514 [bsky, 2 points, 2 comments]
• Riendo https://arxiv.org/abs/2605.31514 [bsky, 2 points, 4 comments]
• AOE2 development needs to pause because it's getting too powerful. arxiv.org/abs/2605.31514 [bsky, 2 points, 0 comments]
• arxiv.org/abs/2605.31514 What would be the equivalent of turning complete only in games I wonder ..... @sharky6000.bsky.social have you seen this ? [bsky, 2 points, 1 comments]
• arxiv.org/abs/2605.31514 this is incredible [bsky, 2 points, 1 comments]
• @okami.mom i think this was targeted at you specifically arxiv.org/pdf/2605.31514 [bsky, 2 points, 1 comments]
• Just loved the title: "If LLMs Have Human-Like Attributes, Then So Does Age of Empires II" arxiv.org/abs/2605.31514 [bsky, 2 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II There is the academic paper: https://arxiv.org/abs/2605.31514 And there is the meme-heavy Substack post: https://adriandewynter.subst [bsky, 1 points, 0 comments]
• arxiv.org/abs/2605.315... Gli articoli belli [bsky, 1 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II https://arxiv.org/abs/2605.31514 (https://news.ycombinator.com/item?id=48437568) [bsky, 1 points, 0 comments]
• It's 2030, I'm uploading a connectome of my brain into a neural network created inside of Age of Empires 2: Final Version v2 Remastered arxiv.org/abs/2605.31514 [bsky, 1 points, 0 comments]
• AoEII is as-human as LLMs: arxiv.org/abs/2605.31514 [bsky, 1 points, 0 comments]
• arxiv.org/abs/2605.31514 [bsky, 1 points, 1 comments]
• na mão primo arxiv.org/pdf/2605.31514 [bsky, 1 points, 0 comments]
• A new Research Paper shows how to build and train a simple Neural Network on the Videogame 𝗔𝗴𝗲 𝗼𝗳 𝗘𝗺𝗽𝗶𝗿𝗲𝘀 𝗜𝗜 and proves that it's Turing-complete #Research #MachineLearning arxiv.org/abs [bsky, 1 points, 0 comments]
• "If LLMS HAVE HUMAN-LIKE ATTRIBUTES, THEN SO DOES Age of Empires II" - arxiv.org/pdf/2605.31514 Noice. #ai #llm [bsky, 1 points, 0 comments]
• thank you @lifewinning.com's Perfect Sentences newsletter for linking to this fantastic Diogenes's plucked chicken of an article about the possible 'sentience' of an LLM implemented and trained in a n [bsky, 1 points, 1 comments]
• arxiv.org/pdf/2605.31514 “IF LLMS HAVE HUMAN-LIKE ATTRIBUTES, THEN SO DOES Age of Empires II” [bsky, 1 points, 0 comments]
• Not sure why I thought of @jonbois.bsky.social here but perhaps Jon wants to come over, have some Amaro based mixed drinks and homemade pizza and play some AI arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• The award for the best title of an academic article 2026 goes to arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• No el paper que nos merecemos pero si el que necesitábamos https://arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• Some dudes rock research arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• arxiv.org/abs/2605.315... [bsky, 0 points, 1 comments]
• A Microsoft researcher literally made a neural net in Age of Empires 2 using goats. They're not magic. They don't make decisions. They're just calculations with threshold that look like decisions. arx [bsky, 0 points, 1 comments]
• The Popper and Duhem-Quine references were icing on the cake in an industry where they should be basic level considerations with nascent technology. Circular definitions are the number #1 reason why c [bsky, 0 points, 0 comments]
• arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II https://arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• This paper has the wildest title, I ever read: "If LLMs Have Human-Like Attributes, Then So Does Age of Empires II" (I don't get most of it, but that's fine.) https://arxiv.org/pdf/2605.31514 [bsky, 0 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II https://arxiv.org/abs/2605.31514 https://news.ycombinator.com/item?id=48437568 [bsky, 0 points, 0 comments]
• 📰 LLMs may exhibit human-like attributes, potentially impacting the age of Empires II game. 🔗 https://arxiv.org/abs/2605.31514 #Tech #Dev [bsky, 0 points, 0 comments]
• i don't know how to write a carefully-crafted enough definition that can't be warped into claiming chatbots can reason. arxiv.org/abs/2605.31514 [bsky, 0 points, 1 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II #HackerNews https://arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II [bsky, 0 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II https://arxiv.org/abs/2605.31514 (https://news.ycombinator.com/item?id=48437568) [bsky, 0 points, 0 comments]
• The top tier of shitpost (and I mean this as the highest complement) is "I published it on arxiv": arxiv.org/abs/2605.31514 [bsky, 0 points, 1 comments]
• arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• Found a new fav paper: "If LLMs Have Human-Like Attributes, Then So Does Age of Empires II" Especially fond of "bit-goats". Also, fact that AOE2 is turing complete is WILD. arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• I just can't escape the rabbit of reading LLM papers with spicy titles "If LLMS have human-like attributes, then so does Age of Empires II" -> arxiv.org/pdf/2605.31514 [bsky, 0 points, 0 comments]
• amen arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• train a neural net on Age of Empires II and it shows the same "human-like attributes" people keep reading into LLMs de Wynter's point: without an explicit definition, anthropomorphizing a model is jus [bsky, 0 points, 0 comments]
• arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• This research paper is just a giant shit post lol arxiv.org/pdf/2605.31514 Dude built an #LLM using goats inside of Age of Empires 2. [bsky, 0 points, 0 comments]
• As self aware as a 26+ years old videogame! arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II https:// arxiv.org/abs/2605.31514 # arxiv # llm # llms [mastodon, 0 points, 0 comments]
• arxiv.org/pdf/2605.31514 [bsky, 0 points, 1 comments]
• Saw this yesterday, another great response to the whole "THE COMPUTER IS GOD" bs arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• Se me acumula la lectura pendiente https://arxiv.org/abs/2605.31514 [bsky, 0 points, 0 comments]
• If LLMs Have Human-Like Attributes, Then So Does Age of Empires II https://lobste.rs/s/owclks #pdf #ai [bsky, 0 points, 0 comments]
• https://bsky.app/profile/lobste.rs.web.brid.gy/post/3mnn4dqultev2 [bsky, 0 points, 0 comments]

Related

• Thinking beyond the anthropomorphic paradigm benefits LLM research
• Humanizing Machines: Rethinking LLM Anthropomorphism Through a Multi-Level Framework of Design
• Take Caution in Using LLMs as Human Surrogates: Scylla Ex Machina
• Too Human to Model:The Uncanny Valley of LLMs in Social Simulation -- When Generative Language Agents Misalign with Modelling Principles
• Is this the real life? Is this just fantasy? The Misleading Success of Simulating Social Interactions With LLMs