We have updated our Privacy Policy and Terms of Use for Eurasia Group and its affiliates, including GZERO Media, to clarify the types of data we collect, how we collect it, how we use data and with whom we share data. By using our website you consent to our Terms and Conditions and Privacy Policy, including the transfer of your personal data to the United States from your country of residence, and our use of cookies described in our Cookie Policy.
Artificial intelligence systems are trained on massive troves of data — but it could use some expert advice. After all, not all data is created equal.
Take the written word: There’s a difference between training on tweets, New York Times articles, classic literature, Wikipedia entries, and academic journals.
AI companies seem to be missing the good stuff, essentially, the expertise. But expertise is often proprietary, sold by companies or behind tight paywalls, and isn’t easy to scrape up for training. One company called Gretel told Fast Company that its platform can be used to anonymize expert data so it can be sold to AI firms.
We’ve already seen a land grab for high-quality training data with AI firms trying to sign licensing deals with news publishers, though some have opted for copyright litigation rather than taking the money. Could we see an AI company buy up a news publisher, or a social media site, or even a publishing house? Facebook parent company Meta reportedly explored buying the publishing giant Simon & Schuster to train its AI systems. A similar acquisition might not be too far in the future.
