Vanishing Gradients Episode 14: Decision Science, MLOps, and Machine Learning Everywhere
23 snips
Nov 20, 2022 
Hugo Bowne-Anderson discusses decision science, MLOps, and the ubiquity of machine learning models. Topics include decision-making under uncertainty, biases in data collection, MLOps and DevOps convergence, digital feedback loops, Google's search evolution, and the impact of modern algorithms on reality perception.
AI Snips
Chapters
Transcript
Episode notes
Separate Decision Quality From Outcomes
- Decisions combine quality and chance; bad outcomes don't imply bad decisions.
- Judge decisions by information available at the time, not by single outcomes.
Report Probabilities Clearly
- Think probabilistically and report uncertainties rather than deterministic statements.
- Use language that reflects chance, e.g., 'one in five chance', to avoid misleading certainty.
Assess Likelihood And Impact Together
- Risk requires assessing both likelihood and impact, not likelihood alone.
- Low probability but high impact events demand different choices than low-impact risks.