Directionally Correct podcast is sponsored by Worklytics! https://www.worklytics.co/directionallycorrect/
Leaked Google Document: We have no moat, but neither does OpenAI: https://simonwillison.net/2023/May/4/no-moat/
ChatGPT Code Interpreter: https://docs.kanaries.net/articles/chatgpt-code-interpreter
Pre-Prints on arxiv.org: https://arxiv.org/
Kelsie’s article: Examining the Use of Game Based Assessment for Hiring Autistic Job Seekers: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8628896/pdf/jintelligence-09-00053.pdf
Results on MTurk are Getting Worse: https://twitter.com/emollick/status/1607589896222576640
ChatGPT as a Reviewer: https://twitter.com/emollick/status/1605022328647077888?s=20
Computers can Never be Held Accountable: https://twitter.com/MIT_CSAIL/status/1484933879710371846?s=20