‘Research often guns for incremental advances, juicing an extra percentage point of accuracy out of a previously proposed model.’
Using AI results for real-world decisions can trigger a range of extra expenses, reports AXIOS. AI labs often concentrate on R&D metrics such as accuracy. Corporate uses should look at possible additional costs associated with:
- Data labelling
- Reliability
- Vulnerability
- Bias
- Explainability
1 big thing: The hidden costs of AI
AXIOS | October 25, 2019 | By Kaveh Waddell