The well for AI training data is running dry. Big Tech heavyweights are taking extraordinary measures to deal with the drought.
AI in Healthcare April 9, 2024
Contrary to popular perception, digital data that’s suitable for training AI models is a finite resource. Bumping up hard against this reality, three of the field’s top players—OpenAI, Google and Meta—have been acting like they’ve had no choice but to potentially cut some ethical corners, skirt their own policies and mull the pros and cons of bending the law.
This is not an opinion. It’s the conclusion of five of the sharpest tech reporters on staff at The New York Times. The nation’s unofficial newspaper of record launched a journalistic investigation after its legal team sued OpenAI and Microsoft last year. At that time the Times accused the companies of using copyrighted news articles to train AI models. The companies...