Understanding AI's Social Thinking Skills

Understanding AI's Social Thinking Skills

Author: Artificially Human September 7, 2025 Duration: 6:25
This podcast was created entirely by AI and is based on the following research paper: Title: How large language models encode theory-of-mind: a study on sparse parameter patternsSource: npj Artificial IntelligenceAuthors: Yuheng Wu et al.Published Date: 2025-08-28Visit www.paper2podcast.com to download the full paper and learn more. Thanks for listening!

Ever find yourself staring at a dense academic paper, wishing you could just absorb its ideas on a walk or during your commute? That’s the space Paper-to-Podcast occupies. Created by Artificially Human, this series takes peer-reviewed research from across the sciences and technology and converts it into an accessible audio format. The unique twist is that the entire production-from the narration to the editorial synthesis-is handled by artificial intelligence. This means each episode is a direct, albeit sometimes unexpectedly quirky, translation of scholarly work into a narrative you can listen to. You’ll encounter breakthroughs in machine learning, surprising studies in behavioral science, and deep dives into emerging tech, all presented with the clarity and occasional charming oddity that AI generation brings. It’s for the eternally curious who love the frontiers of knowledge but appreciate a more fluid, conversational approach than a PDF can offer. Tune in to hear complex concepts unpacked, and to experience firsthand how machine intelligence interprets and communicates human discovery. The result is a consistently fascinating and often wonderfully weird blend of education and experimental audio, making each episode a small adventure in learning.
Author: Language: en-us Episodes: 100

Paper-to-Podcast
Podcast Episodes
Smart Sampling Beats Trained AI [not-audio_url] [/not-audio_url]

Duration: 4:47
This podcast was created entirely by AI and is based on the following research paper: Title: Reasoning with Sampling: Your Base Model is Smarter Than You ThinkSource: arXivAuthors: Aayush Karan, Yilun DuPublished Date: 2…
Evolving Language Models Like Nature [not-audio_url] [/not-audio_url]

Duration: 7:10
This podcast was created entirely by AI and is based on the following research paper: Title: Nature-inspired Population-based Evolution of Large Language ModelsSource: arXivAuthors: Yiqun Zhang et al.Published Date: 2025…
AI as a Teamwork Partner [not-audio_url] [/not-audio_url]

Duration: 6:51
This podcast was created entirely by AI and is based on the following research paper: Title: The Cybernetic Teammate: A Field Experiment on Generative AI Reshaping Teamwork and ExpertiseSource: Harvard Business SchoolAut…
Understanding and Measuring Artificial General Intelligence [not-audio_url] [/not-audio_url]

Duration: 5:36
This podcast was created entirely by AI and is based on the following research paper: Title: A Definition of AGISource: Center for AI SafetyAuthors: Dan Hendrycks et al.Published Date: 2025-10-21Visit www.paper2podcast.c…
Predicting AI Behavior with Model Size [not-audio_url] [/not-audio_url]

Duration: 6:22
This podcast was created entirely by AI and is based on the following research paper: Title: Evaluation Awareness Scales Predictably in Open-Weights Large Language ModelsSource: arXivAuthors: Maheep Chaudhary et al.Publi…
Can AI Models Get Addicted? [not-audio_url] [/not-audio_url]

Duration: 5:16
This podcast was created entirely by AI and is based on the following research paper: Title: Can Large Language Models Develop Gambling Addiction?Source: arXivAuthors: Seungpil Lee et al.Published Date: 2025-09-26Visit w…
Teamwork in AI Language Models [not-audio_url] [/not-audio_url]

Duration: 5:44
This podcast was created entirely by AI and is based on the following research paper: Title: Emergent Coordination in Multi-agent Language ModelsSource: arXivAuthors: Christoph RiedlPublished Date: 2025-10-05Visit www.pa…
Vibe Coding: Fast but Flawed [not-audio_url] [/not-audio_url]

Duration: 5:34
This podcast was created entirely by AI and is based on the following research paper: Title: Vibe Coding in Practice: Motivations, Challenges, and a Future Outlook – a Grey Literature ReviewSource: arXivAuthors: Ahmed Fa…
Fine-Tuning AI Models Can Be Risky [not-audio_url] [/not-audio_url]

Duration: 6:04
This podcast was created entirely by AI and is based on the following research paper: Title: Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend to!Source: arXivAuthors: Xiangyu Qi et al…
Bridging AI and Brain Models [not-audio_url] [/not-audio_url]

Duration: 4:54
This podcast was created entirely by AI and is based on the following research paper: Title: The Dragon Hatchling: The Missing Link Between the Transformer and Models of the BrainSource: arXivAuthors: Adrian Kosowski et…