7 Lessons from building a small-scale AI application
These are good. tl;dr:
- AI programming is stochastic;
- Data quality is real work;
- Models are only as good as the evaluation;
- Trust/Quality is the #1 issue;
- Your training pipeline is your core IP;
- AI is yet another distributed system;
- Don't buy the AI library hype
via Niall Murphy.
#