Brendanjones, 3 months ago

#Docusign have too much juicy data to be left out of the fun:

"Docusign just admitted that they use customer data (i.e., all those contracts, affidavits, and other confidential documents we send them) to train AI"

2/x

https://mastodon.social/@gvwilson/112012277852906749

Brendanjones, 3 months ago

#Tumblr and #Wordpress in on the act:

"Tumblr and WordPress.com are preparing to sell user data to Midjourney and OpenAI"

3/x

https://www.404media.co/tumblr-and-wordpress-to-sell-users-data-to-train-ai-tools/ #OpenAI

otto42, 3 months ago

@Brendanjones https://wordpress.com/blog/2024/02/27/more-control-over-the-content-you-share/

Brendanjones, 3 months ago

@otto42 omfg

"We already discourage AI crawlers from gathering content from WordPress.com and will continue to do so, save for those with which we partner."

Brendanjones, 3 months ago

16 years of user-generated Stack Overflow content is going to be fed into AI to “provide Stack Overflow content directly within Google Cloud”.

4/x

https://meta.stackexchange.com/questions/398127/our-partnership-with-google-and-commitment-to-socially-responsible-ai?cb=1 #AI #StackOverflow #StackExchange

Brendanjones, 3 months ago

Here's one from last year that just came to my attention: #Elsevier have packaged up millions of scientific papers and author profiles for anyone to use for "AI and digital transformation" (their words from https://www.elsevier.com/solutions/datasets).

The complete opposite of #OpenScience.

5/x

https://www.elsevier.com/about/press-releases/elsevier-introduces-authoritative-scientific-datasets-to-fuel-innovation-and and

#Science #AI

Brendanjones, 3 months ago

The dataset includes:

• 19 million full-text articles from peer-reviewed journals
• 17 million author profiles
• 1.8 billion cited references
• 333 million chemical substances and reactions
• 86 million bioactivities and biomedical records
• 35 million chemical patents

Woah. That's humanity's knowledge, for sale if you have the money.