Sabitlenmiş Tweet

Time for a little QA on the AI clause that everyone needs to know when it comes to this AI situation.
I hate to break this to all of you, but if you're concerned about the change to the ToS that added the AI clause, then you fell for rage bait. It's been there for a long time.
Explanation:
Q: "Before you start, can you prove your claim?"
A: The AI clause has been in Twitter's Privacy Policy (not the ToS, the Privacy Policy. They aren't the same thing, but are both binding documents!) since at least September 2023 - because Twitter was on the verge of launching Grok, a generative AI, in November of that year. The previous Privacy Policy is still on the Privacy Policy page, so you can just check it yourself - however, I know some of you will stubbornly refuse to do that, so I have included a screenshot for reference.
Q: "Okay, so then what did the recent change actually do, and why were people angry about it?"
A: The upcoming changes to the ToS moved the AI language out of the Privacy Policy and into the ToS. This was likely done in an effort to consolidate redundant verbiage regarding AI. A handful of people went and looked at this change after it was announced, saw the new wording, and either failed to do their research or knowingly didn't tell you it was in the Privacy Policy for engagement bait. There's no way to tell which one happened but both are bad.
Q: "Fine, but we still can't opt out starting with the new policy!"
A: If you notice, in the screenshot, the policy that has been in place since Grok launched also lacks an opt-out clause despite the fact that an opt-out is offered on the site. This is because you don't need to include a clause to offer an opt-out. This concern came about because a prior draft of the new ToS included a line about private communities, which was removed. What's important to note is that this line was not an opt-out either, but people were already up in arms and emotionally worked up, and were looking for more reasons to be upset. There is no evidence that the opt-out will be removed, but I think we can all agree that Twitter staff should confirm that to their user base anyway. It would help relieve some concerns.
Q: "Okay, but Pirate Software said..."
A: With the verbiage being moved into the ToS, the wording was changed slightly to include the word "generative". Pirate Software seems to think the addition of this term is meaningful, but as the original phrase was launched with Grok (which has always been generative) the change is likely being added simply to ensure compliance with government laws to better define what they have already been doing. There has never been a period where the AI being trained on Twitter has -not- been generative. I personally believe that this is because Pirate Software (like most people) failed to realize that wording was in the privacy policy when he initially discussed it, and spoke on the matter a bit prematurely - then needed a reason to not have to walk his statement back, so he zeroed-in on the "generative" phrasing. Thor is usually extremely thorough, but everyone makes mistakes.
That covers most of the common things I've heard so far. If I get any more information I'll amend it here with a comment. But instead of having hundreds of separate discussions on the topic, I figured it made more sense to just post it once and link to this whenever I see the common misinformation occurring.
English














