Not So Standard Deviations 189 - LLMs and Data Science
14 snips
Aug 5, 2024 
Hilary Parker, a keynote speaker focused on LLMs and data science, joins Roger Peng, an expert in programming languages. They explore the future of programming languages like R in the age of LLMs. Privacy concerns in AI, particularly in email communication, are discussed alongside the balance of honesty in feedback. The conversation highlights the impact of AI on quality control, the enduring legacy of COBOL, and the importance of transparency in data science. They also share humorous insights on the challenges of educating children in complex subjects.
AI Snips
Chapters
Transcript
Episode notes
Bank Statement Example
- Hilary Parker used a bank statement example in her keynote to illustrate LLM uses.
- This sparked privacy concerns among some attendees, highlighting the tension between practicality and privacy.
Email Communication Enhancement
- Use LLMs to improve email communication, especially for sensitive feedback.
- Consider incoming and outgoing filters to ensure professionalism and avoid misinterpretations.
LLMs for Data Preprocessing
- LLMs offer a practical, albeit imperfect, solution for data preprocessing.
- An 80% solution is often sufficient for non-critical analyses, improving speed despite occasional errors.