Language models cannot reliably distinguish  belief from knowledge and fact

Is that stack of scientific papers growing ever higher? Don't you wish you could fill that void in your life when you're off work and actually living your life with more work? Well then Science TLDR is the podcast for you!At some regular interval, a science paper (of my choosing for the time being) is summarized and broken down into a quick 15 minute-ish podcast. My budget is exactly $0 at the moment so the summaries and scripts are written using a variety of LLMs (mostly Gemini and notebookLM, but occasionally ChatGPT 4 and Claud) and read aloud sometimes with a zesty AI voice from 11labs. If I ever have enough money to convert this into something, I am committed to hire real human talent for these jobs.If you have any feedback or suggestions please send them to <a target="_blank" rel="noopener noreferrer nofollow" href="mailto:sciencetldrpod@gmail.com">sciencetldrpod@gmail.com</a>

Science

Is that stack of scientific papers growing ever higher? Don't you wish you could fill that void in your life when you're off work and actually living your life with more work? Well then Science TLDR is the podcast for you! At some regular interval, a science paper (of my choosing for the time being) is summarized and broken down into a quick 15 minute-ish podcast. My budget is exactly $0 at the moment so the summaries and scripts are written using a variety of LLMs (mostly Gemini and notebookLM, but occasionally ChatGPT 4 and Claud) and read aloud sometimes with a zesty AI voice from 11labs. If I ever have enough money to convert this into something, I am committed to hire real human talent for these jobs. If you have any feedback or suggestions please send them to sciencetldrpod@gmail.com

Is that stack of scientific papers growing ever higher? Don't you wish you could fill that void in your life when you're off work and actually living your life with more work? Well then Science TLDR is the podcast for you!At some regular interval, a science paper (of my choosing for the time being) is summarized and broken down into a quick 15 minute-ish podcast. My budget is exactly $0 at the moment so the summaries and scripts are written using a variety of LLMs (mostly Gemini and notebookLM, but occasionally ChatGPT 4 and Claud) and read aloud sometimes with a zesty AI voice from 11labs. If I ever have enough money to convert this into something, I am committed to hire real human talent for these jobs.If you have any feedback or suggestions please send them to <a href="mailto:sciencetldrpod@gmail.com" rel="noopener noreferrer">sciencetldrpod@gmail.com</a>

Share Language models cannot reliably distinguish belief from knowledge and fact

Sign up to save your podcasts

Language models cannot reliably distinguish belief from knowledge and fact

Language models cannot reliably distinguish belief from knowledge and fact