Elon Musk's updated new chatbot Grok-2 shows how badly AI can...
The internet is being flooded with scandalous images thanks to the rollout of new artificial intelligence, sparking major concerns. Artificial intelligence (AI) is a powerful new force that has been unleashed on the world. But with great power comes great responsibility.
The latest iteration of AI chatbot Grok shows how badly things can go wrong. Released on August 13, Grok-2 is now allowing X Premium users to generate images based on prompts.
This tech has been around for a while, with Meta’s Llama 3 and Google’s Gemini joining Open AI’s GPT-4 as the latest offerings from the big players.
Stricter Content Guidelines
However, most firms have put in place stringent content guidelines, with restrictions relating to illegal, hateful, and sexually explicit or violent content. However X (formerly Twitter) users are reporting that the guidelines are failing on Grok, which was developed by free speech "warrior" Elon Musk.
Ostensibly, Grok will not generate "images that could be used to deceive or harm others, like deepfakes intended to mislead, or images that could lead to real-world harm", a spokesman told The Verge.
Concerns Over Outrageous Images
When the updated Grok was released on Tuesday, though, it was immediately apparent that these guardrails were not working. Instead, Grok-2 allowed users to generate various outrageous images.
Pictures posted by users included Donald Trump appearing as an ISIS terrorist, Joe Biden in a bikini and Elon Musk as a school shooter. Some of the images are too graphic to publish.
“Grok 2.0 is completely unhinged,” one user wrote. “It’s been one day, and everyone on X is going nuts over its uncensored image capabilities.”
Responses to Grok-2
Alejandra Caraballo, a clinical instructor at the Harvard Law School Cyberlaw Clinic, wrote: “Oh my god. Grok has absolutely no filters for its image generation. This is one of the most reckless and irresponsible AI implementations I’ve ever seen.”
Other chatbots, like OpenAI’s Dall-E 3, will not create images of real people. Some users expressed concern about the risks generative AI might pose in the US presidential election in November specifically, and for the spread of misinformation generally.
Regulatory Scrutiny
Mr. Musk himself has posted manipulated content on X, including a digitally-altered video of Vice President Harris, apparently in breach of his own website’s guidelines.
It comes as regulators increasingly put X under the microscope. In July, the Office of the Australian Information Commissioner warned X that it may be in breach of the nation’s privacy laws for harvesting data without informing users.
Investigations and Protections
In December last year, the European Commission revealed it was investigating X for potential breaches to the Digital Services Act linked to “risk management, content moderation, dark patterns, advertising transparency and data access for researchers”.
X has various protections in the US including section 230 of the Communications Decency Act, which shields tech companies from liability over user content. However, US authorities are looking at cracking down on AI impersonation, deepfakes, and misinformation.
Elon Musk's Response
Elon Musk, the world’s richest person, was unconcerned by the Grok controversy and said he wanted people to “have some fun”. Grok uses FLUX.1 by Black Forest Labs to generate images rather than an in-house system.
Mr. Musk wrote: “We have our own image generation system under development, but it’s a few months away, so this seemed like a good intermediate step for people to have some fun.”
In response to a clip of a Donald Trump-style character holding a woman resembling Elon Musk, the South African-born billionaire wrote: “Well if I live by the sword, I should die by the sword.”
With the rising cost of living stretching budgets further than ever, cost-conscious Aussies are cutting back. It’s left one key industry in a battle to survive. Amazon executives are being grilled on the company’s murky development of AI and how audio recordings from inside homes are used.
A highly anticipated report on social media’s effects on Australian society has been delayed after a ministerial shakeup.
