Share this article on:
Breaking news and updates daily. Subscribe to our Newsletter
Earlier this month, X (formerly Twitter) chief executive Elon Musk introduced the world to Grok, the social media company’s new AI chatbot, designed to take on the likes of Google Bard and OpenAI’s ChatGPT.
The new bot, which derives its name from Robert A. Heinlein’s novel Stranger in a Strange Land and has been modelled after Hitchiker’s Guide to the Galaxy in classic nerdy Musk fashion, is advertised as a chatbot with “a bit of wit and … a rebellious streak.”
While based on a large language model (LLM) just like its rivals, its real party trick is its access to posts on X.
“It will also answer spicy questions that are rejected by most other AI systems,” wrote X on its own site.
Announcing Grok!— xAI (@xai) November 5, 2023
Grok is an AI modeled after the Hitchhiker’s Guide to the Galaxy, so intended to answer almost anything and, far harder, even suggest what questions to ask!
Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use…
X has faced controversy for its moderation policies, with users questioning whether it is a platform for people to say whatever they please or a hot pot of hate speech.
As a result, Grok may end up spreading misinformation in the same way as X does.
As RMIT University’s Dr Nataliya Ilyushina points out, LLMs “often reflect biases as they are trained by data generated by humans”.
However, Grok’s ability to draw on X posts in real time, plus its “rebellious” attitude, may lead it to spread misinformation even more than other LLM-powered chatbots.
“There has been considerable debate surrounding [X’s] moderation policies and controversies regarding its status as either a platform for free speech or hate speech,” said Ilyushina.
“It is possible that Grok, which will train itself using X’s posts, will inherit these issues and be subject to the same attacks concerning the limits of free speech.
“It could also raise concerns that those who control the moderation on X will influence the direction of the ‘bias’ in this new AI language model. “
Musk warned users who are “intolerant of the brand of human it may present” to steer clear of the chatbot; however, the tech tycoon may find that some individuals may have no choice but to avoid Grok.
“With X’s challenges related to misinformation and the ongoing development of the Misinformation and Disinformation Bill in Australia, the platform could potentially face censorship in Australia,” added Ilyushina.
X has landed itself in hot water in Australia of late, most recently for failing to pay fines related to a lack of social media moderation and not responding to questions relating to child sexual abuse material on the platform.
The $610,500 fine was issued by the Office of the eSafety Commissioner last month and was given 28 days to pay. However, with that deadline passing yesterday (15 November), X has still failed to pay up.
“Twitter/X has not paid the infringement notice within the allotted time frame, and eSafety is now considering further steps,” an eSafety spokesperson said in a widely reported statement.
“If X does not pay the fine or does not comply with the request for information, the eSafety Commissioner could seek a civil penalty from the Federal Court, which could lead to daily fines of up to $782,000, backdated to February 2023, when the first notice was issued. It could add up to millions of dollars.”
Comments powered by CComment