Content-type: text/html Downes.ca ~ Stephen's Web ~ The OpenAI Endgame

Stephen Downes

Knowledge, Learning, Community

This is a reasonably intelligent discussion about the potential outcomes of the court case between the New York Times and OpenAI on the use of content as training data. Mike Loukides makes an interesting prediction: "OpenAI will settle with The New York Times out of court, and we won't get a ruling." This creates a good thought experiment because, as he says, such a settlement "will set a de-facto price on training data. And that price will no doubt be high... sufficiently high enough to deter OpenAI's competitors." It also keeps the pool of those who provide training data pretty small. "This is chilling: if all AI applications go through one of a small group of monopolists, can we trust those monopolists to deal honestly with issues of bias?" Ultimately, the argument here is in favour of open data supporting open AI, but if a settlement were recognized as de facto law this would be that much more difficult to achieve.

Today: 0 Total: 1477 [Direct link] [Share]


Stephen Downes Stephen Downes, Casselman, Canada
stephen@downes.ca

Copyright 2024
Last Updated: May 22, 2024 5:44 p.m.

Canadian Flag Creative Commons License.

Force:yes