# AI Finds A Way arXiv: https://arxiv.org/abs/2608.23875 ยท HTML: https://mlsanity.com/paper/2608.23875/ ## Key facts - Cited by 3 papers in the map (data through 14 Oct 2026). - Builds on 68 of its 256 references that are in the map. - First posted on arXiv 24 Aug 2026. ## Builds on (most cited first, 30 of 68) - [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023) - [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014) - [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017) - [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024) - [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021) - [Auto-Encoding Variational Bayes](https://mlsanity.com/paper/1312.6114/) (2013) - [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025) - [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022) - [Sparks of Artificial General Intelligence: Early experiments with GPT-4](https://mlsanity.com/paper/2303.12712/) (2023) - [Emergent Abilities of Large Language Models](https://mlsanity.com/paper/2206.07682/) (2022) - [Do As I Can, Not As I Say: Grounding Language in Robotic Affordances](https://mlsanity.com/paper/2204.01691/) (2022) - [Language Models (Mostly) Know What They Know](https://mlsanity.com/paper/2207.05221/) (2022) - [Concrete Problems in AI Safety](https://mlsanity.com/paper/1606.06565/) (2016) - [OpenAI o1 System Card](https://mlsanity.com/paper/2412.16720/) (2024) - [Mastering Diverse Domains through World Models](https://mlsanity.com/paper/2301.04104/) (2023) - [Code as Policies: Language Model Programs for Embodied Control](https://mlsanity.com/paper/2209.07753/) (2022) - [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023) - [Dota 2 with Large Scale Deep Reinforcement Learning](https://mlsanity.com/paper/1912.06680/) (2019) - [The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery](https://mlsanity.com/paper/2408.06292/) (2024) - [Towards Understanding Sycophancy in Language Models](https://mlsanity.com/paper/2310.13548/) (2023) - [DeepSeek-V3 Technical Report](https://mlsanity.com/paper/2412.19437/) (2024) - [AlphaEvolve: A coding agent for scientific and algorithmic discovery](https://mlsanity.com/paper/2506.13131/) (2025) - [Measuring Faithfulness in Chain-of-Thought Reasoning](https://mlsanity.com/paper/2307.13702/) (2023) - [The StarCraft Multi-Agent Challenge](https://mlsanity.com/paper/1902.04043/) (2019) - [Generating Sequences With Recurrent Neural Networks](https://mlsanity.com/paper/1308.0850/) (2013) - [Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training](https://mlsanity.com/paper/2401.05566/) (2024) - [Model-Based Reinforcement Learning for Atari](https://mlsanity.com/paper/1903.00374/) (2019) - [Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision](https://mlsanity.com/paper/2312.09390/) (2023) - [Scalable agent alignment via reward modeling: a research direction](https://mlsanity.com/paper/1811.07871/) (2018) - [The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search](https://mlsanity.com/paper/2504.08066/) (2025) ## Built on by (most cited first, 3 of 3) - [A Case Study on Emergent Cheating and Whistleblowing in Autonomous Research Swarms](https://mlsanity.com/paper/2609.04170/) (2026) - [Directing large language models to follow the letter or spirit of the law](https://mlsanity.com/paper/2609.23083/) (2026) - [Hyper Algorithm Design Agent: Evolving Learnable Optimizer from Zero](https://mlsanity.com/paper/2609.35328/) (2026) ## Abstract (arXiv) Artificial Intelligence (AI) algorithms frequently learn creative and unexpected solutions, surprising even expert researchers who develop and study them. They often astonish practitioners by discovering unanticipated behavior, exploiting loopholes in reward signals, or spontaneously uncovering previously unknown scientific phenomena. However, accounts of such unconventional behavior across machine learning are seldom formally documented. This work presents 26 curated firsthand anecdotes from various machine learning subfields representing the work of over 100 researchers. These anecdotes showcase the capability of modern AI systems to circumvent human-imposed design limitations and discover unexpected solutions to the tasks we train them on. Furthermore, these accounts are particularly important for the safety of future AI systems. They illustrate the fundamental challenge of aligning models with human values without diminishing their creativity, so they can make surprising discoveries without producing surprising, potentially harmful outcomes. The paper first details AI achieving superhuman success through reinforcement learning across many challenging domains. However, reward-driven optimization can fail when the model learns to hack an underspecified reward or unarticulated constraint. We then present case studies suggesting that harnessing internet-scale foundation models (FMs) has not resolved these fundamental challenges and, in fact, can supercharge them. Nevertheless, we argue that these same learning dynamics can be harnessed to accelerate scientific discovery. Finally, we hope this work provides a consolidated resource to inform future research and demonstrates that the tendency toward unexpected behaviors is commonplace in modern AI, highlighting the need to anticipate and manage AI's capacity for innovative, yet unpredictable, solutions. (abstract abridged)