AI Applied: Covering AI News, Interviews and Tools - ChatGPT, Midjourney, Gemini, OpenAI, Anthropic ChatGPT Almost Job Ready, But Claude is Closer?!
13 snips
Oct 10, 2025 Conor and Jaeden dive into the impact of AI on the workforce, comparing tools like ChatGPT and Claude. They discuss how AI excels at specific tasks but what that means for job security. The importance of trust and transparency in AI evaluations takes center stage. The hosts emphasize that user experience can be more crucial than raw performance. Additionally, they explore how businesses can strategically select AI models based on industry-specific performance metrics. Tune in for insights on leveraging AI for career advancement!
AI Snips
Chapters
Transcript
Episode notes
Benchmarks Show Task-Level Progress
- OpenAI ran a GDP Val benchmark to compare AI outputs against experienced professionals across industries.
- Benchmarks moving from tests to real task evaluations reveal how models approach job functions.
Task Evaluations Reveal 'Street Smarts'
- Evaluating models on real tasks measures 'street smarts' not just book knowledge.
- Task-based measures reveal practical capabilities relevant to enterprise use.
Tasks, Not Whole Jobs, Are Most Vulnerable
- LLMs excel at discrete, repeatable tasks but not yet full jobs composed of diverse responsibilities.
- Roles made of a few core tasks (e.g., customer service, paralegals) face higher automation risk.