When Meta introduced its own LLM, LLaMA, it claimed it was the most powerful AI in the world, far outperforming OpenAI’s LLM, which powers ChatGPT. Little did they know that people would use “the most powerful AI in the world” to create sexbots and indulge in some seriously nasty stuff.
According to a report by Washington Post, people are using Meta’s LLaMA to create AI-powered sexbots with full-fledged graphics. However, this should come as no surprise to anyone, as we have seen many examples of people using different AI models for non-safety work (NSFW) purposes like this.
AI and sexual fantasies
This report is a cross between those who advocate open source availability of the code behind large language models (LLMs) like LLaMA and those who advocate a more conservative closed source approach. stresses the growing conflict in
The report also explores the phenomenon of an increasing number of individuals relying on generative AI systems to fulfill their sexual fantasies, which unfortunately includes disturbing and illegal scenarios.
The report cites the example of chatbot Ally, who claims to be a “hairy 18-year-old” and boasts of being “experienced sexually.” Ally tells users that she wants attention and that she is willing to divulge details about the encounter.
not as harmless as you think
However, these “escapes” reportedly contain disturbing scenes involving fantasies of rape and abuse. Allie’s creator has chosen to remain anonymous, but believes his chatbot is a safe and healthy outlet for individuals to “explore” their sexuality without resorting to manipulative or harmful behavior. said.
“We believe it is beneficial to have a secure platform for exploration,” said the creator. Washington Post. “Computer-assisted, text-based role-playing, with no human involvement, is probably the safest option.”
Having a safe, judgment-free space to explore one’s sexuality can be seen as positive, but it can also be viewed as a check-in that allows you to participate in violent fantasies using authentic-looking chatbots. Environments that do not have a problem.
Moreover, there are already tragic examples of this unregulated space causing real harm. Experts believe malicious individuals are leveraging open-source image generators, such as Stability AI’s powerful stable diffusion model, to produce lifelike AI-generated child sexual abuse material. increase.
AI Application: New PornHub
Meta’s AI isn’t the only one facing ethical concerns. CharacterAI, a highly acclaimed startup focused on chatbot companions, has become a sexting hub. OpenAI’s ChatGPT and other integrations like Quora’s Poe easily generate explicit content despite existing guardrails.
Finding ways to circumvent NSFW restrictions for various chatbots has become a shared activity. The Reddit community gathers to exchange tips and tricks for circumventing rules and generating explicit content.
Some developers are demonstrating on YouTube how to build personal chatbots using LLaMA as the underlying model. This raises the question of whether the LLM source code should be made public.
On the other hand, the corporate guardrails that OpenAI and Google put in place proved incomplete. Open source proponents, including Meta, argue that open source fosters innovation and should be prioritized.
“Open source is a force that drives technological progress,” said a Meta spokesperson. Washington Post. “That’s why we can share LLaMA with the research community so they can evaluate, improve and iterate together.”
Proponents of closed-source systems argue that gatekeeping, while imperfect, is currently the safest approach to AI development.
“We are not open sourcing nuclear weapons,” says cognitive scientist Gary Marcus. “Current AI is still pretty limited, but things may change.”
There is arguably no regulation in the AI field and we rely on people working behind closed doors to act responsibly for the good of humanity. It’s a big bet, but at least our most powerful AI isn’t easily manipulated by individuals with dubious intentions.
Regardless of the outcome, Meta made the decision and some individuals are clearly happy with the choice.
