Why it matters
Performance depends on objectives, training data, human labelling, computation and evaluation, not autonomous understanding. Uses in science and public life bring benefits alongside error, bias, opacity, labour displacement, privacy, security, concentration and environmental costs.
- Date
- 2012–2023 CE · selected deep-learning adoption, not the origin of artificial intelligence
- Historical setting
- Digital & Global Age · Global
- People & communities
- statisticians and computer scientists · neural-network researchers · hardware and infrastructure teams · data creators and labelers · domain experts · affected communities · auditors and regulators



