Transforming Agency. On the mode of existence of Large Language Models

Abstract

This paper investigates the ontological characterization of Large Language Models (LLMs) like ChatGPT. Between inflationary and deflationary accounts, we pay special attention to their status as agents. This requires explaining in detail the architecture, processing, and training procedures that enable LLMs to display their capacities, and the extensions used to turn LLMs into agent-like systems. After a systematic analysis we conclude that a LLM fails to meet necessary and sufficient conditions for autonomous agency in the light of embodied theories of mind: the individuality condition (it is not the product of its own activity, it is not even directly affected by it), the normativity condition (it does not generate its own norms or goals), and, partially the interactional asymmetry condition (it is not the origin and sustained source of its interaction with the environment). If not agents, then ... what are LLMs? We argue that ChatGPT should be characterized as an interlocutor or linguistic automaton, a library-that-talks, devoid of (autonomous) agency, but capable to engage performatively on non-purposeful yet purpose-structured and purpose-bounded tasks. When interacting with humans, a "ghostly" component of the human-machine interaction makes it possible to enact genuine conversational experiences with LLMs. Despite their lack of sensorimotor and biological embodiment, LLMs textual embodiment (the training corpus) and resource-hungry computational embodiment, significantly transform existing forms of human agency. Beyond assisted and extended agency, the LLM-human coupling can produce midtended forms of agency, closer to the production of intentional agency than to the extended instrumentality of any previous technologies.

References in corpus (45)

• An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
• Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
• Sparks of Artificial General Intelligence: Early experiments with GPT-4
• Scaling Laws for Neural Language Models
• Gemini: A Family of Highly Capable Multimodal Models
• Adapted Large Language Models Can Outperform Medical Experts in Clinical Text Summarization
• Tree of Thoughts: Deliberate Problem Solving with Large Language Models
• Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language models
• Toolformer: Language Models Can Teach Themselves to Use Tools
• The Debate Over Understanding in AI's Large Language Models
• Constitutional AI: Harmlessness from AI Feedback
• How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation, and Detection
• Measuring Massive Multitask Language Understanding
• RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control
• Self-Refine: Iterative Refinement with Self-Feedback
• Language Models (Mostly) Know What They Know
• Augmented Language Models: a Survey
• Mixtral of Experts
• Qwen2.5 Technical Report
• Open X-Embodiment: Robotic Learning Datasets and RT-X Models
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• From task structures to world models: What do LLMs know?
• Faith and Fate: Limits of Transformers on Compositionality
• AgentBench: Evaluating LLMs as Agents
• On the Planning Abilities of Large Language Models : A Critical Investigation
• Understanding the planning of LLM agents: A survey
• Large Language Models as General Pattern Machines
• Let's Verify Step by Step
• Language Models Meet World Models: Embodied Experiences Enhance Language Models
• More Agents Is All You Need
• People cannot distinguish GPT-4 from a human in a Turing test
• Agents: An Open-source Framework for Autonomous Language Agents
• Q-Transformer: Scalable Offline Reinforcement Learning via Autoregressive Q-Functions
• Generative midtended cognition and Artificial Intelligence. Thinging with thinging things
• Language Models as Agent Models
• Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models
• RoboCat: A Self-Improving Generalist Agent for Robotic Manipulation
• OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments
• Do Androids Laugh at Electric Sheep? Humor "Understanding" Benchmarks from The New Yorker Caption Contest
• MuSR: Testing the Limits of Chain-of-thought with Multistep Soft Reasoning
• From Mind to Machine: The Rise of Manus AI as a Fully Autonomous Digital Agent
• MLLM-Bench: Evaluating Multimodal LLMs with Per-sample Criteria
• S-Agents: Self-organizing Agents in Open-ended Environments
• Post Turing: Mapping the landscape of LLM Evaluation
• Do language models plan ahead for future tokens?