Sdorn Provides Timely and Accurate Technology News, Covering APP, AI, IoT, Cybersecurity, Startup and Innovation.
⎯ 《 Sdorn • Com 》

Artists Drop Twitter Over Elon Musk's Plan to Train His AI Project on Tweets

2023-08-02 05:52
Elon Musk's plan to tap user-generated content on Twitter to train his xAI startup is
Artists Drop Twitter Over Elon Musk's Plan to Train His AI Project on Tweets

Elon Musk's plan to tap user-generated content on Twitter to train his xAI startup is not sitting well with artists who are now threatening to stop posting on platform.

"We will use the public tweets—obviously not anything private—for training...just like everyone else has,” Musk said during a Twitter Spaces event about xAI last month.

Musk’s comment seemed to be directed more at creating a text-based chatbot. Nevertheless, a growing number of artists are concerned xAI will also apply to image generation and eventually become smart enough to copy their art styles.

"I love y’all, thanks for years of support, but I’ll be deleting my Twitter in the next coming weeks because of this,” tweeted Nicole Rifkin, an artist who’s done projects for several major publications, including The New Yorker.

Another artist, Yingjue Chen, who’s also a production designer for Netflix, chimed in and tweeted: “Lol. No thank you. Will no longer post art or client work.”

Meanwhile, character designer Max Ulichney, posted: “I'll be deleting all of my art from here very soon. I'm deeply sad about it. Twitter was my favorite site and the art community here helped make me the artist I am today. It was the word of mouth here that launched MaxPacks. I'll be forever grateful. Find me everywhere else.”

Twitter data is clearly valuable for AI training since posts are often fresh, filled with up-to-date information, and cover a wide range of topics, language styles, and original content. But does scraping that information amount to collecting people’s data without their consent? A growing number of groups, including authors, have sued AI companies, noting the same programs could one day plagiarize their writing style or work for future projects.

The Federal Trade Commission is also investigating OpenAI, the maker of ChatGPT and DALL-E 2, over how it collects data to train the company’s programs. So it’s possible regulators will one day intervene to set up clear privacy rules on AI training data.

In the meantime, Musk’s plan to tap Twitter’s data for AI training is also turning off writers. “Could this loser creep stop fucking with millions of livelihoods for one fucking second,” tweeted horror author Gretchen Felker-Martin.