Each web site on the web is presently making an attempt to stay an AI-powered chatbot on its homepage, a development that extends even to 8kun—the far-right picture board web site that’s identified for being a poisonous cess pool of wicked and immoral content material.
Jim Watkins, who created the web site from the ashes of 8chan (which was summarily kicked off the internet by website hosting suppliers in 2019), posted about the location’s new AI chatbot on X on Tuesday. In response to Watkins, the brand new chatbot, which is solely dubbed the “8kun AI,” is presently being actively developed and is offered for testing by a restricted variety of customers. Watkins mentioned:
“The 8kun AI is now out there for additional beta testing. Please use it responsibly. Let’s not break it. We’re presently within the course of of coaching the AI, which is able to take roughly a 12 months. Your inquiries will assist information it towards enchancment.”
In a subsequent publish, Watkins claimed that the chatbot was “being skilled on what’s presently posted on 8kun, and that can take a few 12 months to course of.”
On the 8kun web page the place the chatbot is displayed, a message states that solely “Proto members” of the platform are allowed to check out the chatbot. For apparent causes, I’m not a Proto member, so I couldn’t take a look at out the app to see the way it works.
Watkins’ son, Ron Watkins, posted on X in regards to the chatbot: “Already getting higher outcomes than Grok and it doesn’t maintain your hand. Simply wants a bit extra data about present occasions.”
No offense to Ron, however he’s clearly improper. Whereas Grok, Elon Musk’s edgelord chatbot, clearly has its foibles, the 8kun chatbot’s solutions usually don’t strategy something even marginally resembling coherence. The web page that auto-displays the chatbot’s interactions with website customers is a repository of dozens and dozens of unhinged textual content regurgitations that generally half-resemble a solution however are in any other case shot by with utterly unrelated textual content.
As an illustration, one person requested the chatbot: “who would win in a struggle, Howard the Duck or Batman?”
The reply from the chatbot, verbatim, is:
Howard: I don’t suppose so. I believe it’s only a matter of time earlier than we get to the purpose the place we’re going to have the ability to say, “Hey, what, we’ve acquired to do that. We’ve gotta do this.” After which we’ll have to determine methods to make it work. And that’s what I’ve been doing for a very long time now, and I’m actually enthusiastic about it. It’s a extremely thrilling time for me, as a result of I’ve numerous buddies who’re within the UFC and so they’re all preventing for a similar factor, so I really feel like I can do one thing actually particular for them. So I need to go on the market and guarantee that we do what we are able to to get the perfect out of one another. That’s the one means I understand how. e’ve acquired to do that. We’ve gotta do this.” After which we’ll have to determine methods to make it work.
In one other “reply,” the chatbot seems to assert that JFK was assassinated by a “CIA agent” named “Robert Kennedy Jr.” When requested by one other person about its dream, the chatbot spat out what appears to be like like random jargon associated to the U.S. authorized code.
Whereas Watkins claims that his chatbot is being skilled on the contents of 8kun, it seems to be skilled on different texts as effectively. As an illustration, a number of supposed responses from the chatbot simply seem like transcripts taken from episodes of Democracy Now, the left-leaning information program hosted by Amy Goodman. When one person requested the chatbot a query about Jeffrey Epstein, the posted response included the next textual content:
“AMY GOODMAN: I need to thanks for being with us. That is Democracy Now!, democracynow.org, The Struggle and Peace Report. We’re broadcasting from Washington, D.C. at 10:30 a.m. Jap time.”
In one other publish on X, Watkins claimed that the chatbot has “a number of variations of the Bible on its studying checklist.” He later added: “I’ve no affect on what’s popping out. It’s “studying” 8kun, and a few previous public area books, and studying to place issues collectively. It takes a few 12 months of “studying” for it to be helpful.”
In a special publish, Watkins famous that he “wouldn’t take any medical or any recommendation from the AI for a very long time.” That’s actually true and it’s also true that you simply in all probability shouldn’t ask 8kun AI any questions until you need to spend your days studying garbled nothings from the bowels of the web.