caption
a screenshot of the text:
Tech companies argued in comments on the website that the way their models ingested creative content was innovative and legal. The venture capital firm Andreessen Horowitz, which has several investments in A.I. start-ups, warned in its comments that any slowdown for A.I. companies in consuming content “would upset at least a decade’s worth of investment-backed expectations that were premised on the current understanding of the scope of copyright protection in this country.”
underneath the screenshot is the “Oh no! Anyway” meme, featuring two pictures of Jeremy Clarkson saying “Oh no!” and “Anyway”
screenshot (copied from this mastodon post) is of a paragraph of the NYT article “The Sleepy Copyright Office in the Middle of a High-Stakes Clash Over A.I.”
Current large model would all be sued to death, no license with IP owner yet, would kill all existing commercial large models. Except all IP owner are named and license granted retroactive, but sound unlikely.
Hundred of IP owner company and billion of individual IP owner setting prices will probably behave like streaming: price increase and endless fragmentation. Need a license for every IP owner, paperwork will be extremely massive. License might change, expire, same problem as streaming but every time license expire need to retrain entire model (or you infringe because model keep using data).
And in the EU you have right to be forgotten, so excluded from models (because in this case not transformative enough, ianal but sound like it count as storing), so every time someone want to be excluded, retrain entire model.
Do not see where it possible to create large model like this with any amount of money, time, electricity. Maybe some smaller models. Maybe just more specific for one task.
Also piracy exists, do not care about copyright, will just train and maybe even open source (torrent). Might get caught, might not, might become dark market, idk. Will exist though, like deepfakes.