Artificial intelligence (AI) promises to transform businesses and life in general, although it does face five interconnected limitations:
- Data labeling: Humans must label and categorize AI’s underlying data, which can be a sizable and error-prone chore.
- Obtaining massive training data sets: The current wave of machine learning requires users to not only label training data sets, but also ensure they are sufficiently large and comprehensive. Massive data sets can be difficult to obtain or create for many business use cases.
- The “explainability” problem: Larger and more complex models make it hard to explain, in human terms, why AI reached a certain decision (and even harder when it reaches it in real time).
- Generalizability of learning: Unlike the way humans learn, AI models have difficulty carrying their experiences from one set of circumstances to another.
- Bias in data and algorithms: Negative consequences can include misinformed recruiting decisions, misrepresented scientific or medical prognoses, distorted financial models and criminal-justice decisions, and misapplied legal issues.