Technology

37705 readers

356 users here now

A nice place to discuss rumors, happenings, innovations, and challenges in the technology sphere. We also welcome discussions on the intersections of technology and society. If it’s technological news or discussion of technology, it probably belongs here.

Remember the overriding ethos on Beehaw: Be(e) Nice. Each user you encounter here is a person, and should be treated with kindness (even if they’re wrong, or use a Linux distro you don’t like). Personal attacks will not be tolerated.

Subcommunities on Beehaw:

This community's icon was made by Aaron Schneider, under the CC-BY-NC-SA 4.0 license.

founded 2 years ago

MODERATORS

[email protected]

268

Reddit signs $60M contract allowing AI company to train its models on the social media platform's content (www.reuters.com)

submitted 8 months ago by [email protected] to c/[email protected]

110 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[–] [email protected] 11 points 8 months ago (2 children)

$60m doesn't seem like that much in an era where twitter could (have been) sold for $40b.

[–] [email protected] 4 points 8 months ago (1 children)

60 million a year for access to the relatively public data... That seems pretty good to me tbh.

[–] [email protected] 1 points 8 months ago (1 children)

Maybe, but with people are saying reddit's main value proposition is access to AI training data, and that reddit is worth n billion dollars, $60m seems like a pittance.

[–] [email protected] 1 points 8 months ago (1 children)

Its just an API, right?

[–] [email protected] 3 points 8 months ago (1 children)

No, it's really not.

Firstly, while the data may be public, it's not "free". Scraping reddit and using it to train an AI would likely contravene their terms of use, you'd end up facing similar copyright issues that the current generation of bots has.

Secondly, scraped data would be incomplete, you wouldn't get anything edited or "deleted", which would surely be available if you paid them. The edits and deletes would be very valuable for AI training.

Thirdly, you would get the meta that reddit has. Geolocation, user agent, alt accounts, browsing habits, et cetera.

Fourthly, you wouldn't get exclusivity. Locking out a competitor is worth something.

[–] [email protected] 1 points 8 months ago (1 children)

Idk why you are talking about scraping when I said API?

And is all that information in the training contract?

[–] [email protected] 2 points 8 months ago

I assumed that when you said "it's just an API" you were saying you're paying $60m for an API as opposed to scraping for free.

Is all what information in the training contract?

[–] [email protected] 1 points 8 months ago* (last edited 8 months ago)

$44B was a bad deal, good luck looking for another Elon Musk 😜