Posted 4/6/2024, 10:16:00 PM
Tech Giants Use Questionable Methods to Obtain Massive Amounts of Data for AI Models
- AI models need huge amounts of data to improve, possibly more than what's publicly available by 2026
- OpenAI, Google, and Meta have used questionable methods to get more data like scraping YouTube, Google Docs, Maps reviews without permission
- This raises serious concerns about privacy and copyright as people's works are used without consent or compensation
- Some Google employees knew about OpenAI's scraping but didn't report it as Google did similar things
- A Meta exec meeting showed no concerns about using copyrighted material for AI training without paying creators