Over winter break, while trying to open the social media app Bluesky on my phone, I accidentally opened X instead. On X, I noticed notifications from verified users forced onto users, regardless of whether or not you follow them.
Out of curiosity, I wanted to see what kind of slop was being peddled to me, and at first glance it looked normal. Most of the posts were of people posting various selfies – nothing unheard of for social media. But when I looked at the comments, they caught me way off guard. “Grok, put this person in a cosplay of [this character].” “Grok, put this person in a bikini.” “Grok, put this person in a translucent micro-bikini.” It was almost comedic, in a shocking sort of way, how some people were pushing for Grok – X’s inhouse artificial intelligence (AI) – to generate pornographic content. Nowhere in these comments was consent asked for or given. These people were just doing this out in the open without hesitation.
And yet, somehow it gets worse. There are instances of people asking Grok to manipulate photos of children to become sexual, something that is sickening to think about. This is the product of Elon Musk placing such an emphasis on being resistant to supposed censorship on X. I can only hope that laws can be made to ban this use of AI.
There is another thing that really frustrates me about all of this: the news coverage of this topic, and the pattern found in their headlines. “Grok says this,” “Grok apologizes for that,” “Grok did this.” They all refer to Grok as a person, as something capable of thinking and acting for itself.
Grok did not do any of these things. It is an AI, or more accurately, a large language model (LLM). It can neither act nor think independently; it can only respond to input and provide approximations based on all the data it has been trained on to give a relevant and acceptable answer. It can help to think of it as providing an average of a series of numbers but applied to text.
Why do we continue to act like Grok is a person? I do not want to be unfair here, and I understand how easy it is to fall into the habit of personifying chatbots like that. One can have something like a conversation with it, and it has the appearance of having a personality. These make it very easy to refer to it as its own being, since it certainly seems to act like it has its own thoughts and feelings. I would even go so far as to say it could fool a few people administering the Turing Test against it. It is for these reasons that it is so important to make the clarification that AI is not a person. Grok is an LLM that cannot be held accountable for its actions. It is important to understand that because it had to be given a command to do something it did not act on its own accord.
It leaves me frustrated and disappointed that well-known news publications, such as The New York Times, Reuters and Wired, are holding Grok accountable rather than the company who created it or the people asking it to make pornographic content. We should be demanding answers from Elon Musk and X.AI Corp., not Grok. Look at the people who opted not to put safeguards in place instead, the people who allowed Grok to make these kinds of image edits.
