Each web site on the web is at present attempting to stay an AI-powered chatbot on its homepage, a pattern that extends even to 8kun—the far-right image board website that’s recognized 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 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 just dubbed the “8kun AI,” is at present being actively developed and is on the market 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 at present 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 submit, Watkins claimed that the chatbot was “being skilled on what’s at present 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 information about present occasions.”
No offense to Ron, however he’s clearly unsuitable. Whereas Grok, Elon Musk’s edgelord chatbot, clearly has its foibles, the 8kun chatbot’s solutions usually don’t method 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 typically half-resemble a solution however are in any other case shot by with utterly unrelated textual content.
For example, one consumer 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 assume so. I feel 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, you recognize what, we’ve received to do that. We’ve gotta do this.” After which we’ll have to determine how one can 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 a number of pals who’re within the UFC they usually’re all preventing for a similar factor, so I really feel like I can do one thing actually particular for them. So I wish to go on the market and ensure that we do what we will to get the most effective out of one another. That’s the one manner I understand how. e’ve received to do that. We’ve gotta do this.” After which we’ll have to determine how one can make it work.
In one other “reply,” the chatbot seems to say that JFK was assassinated by a “CIA agent” named “Robert Kennedy Jr” (you recognize, the President’s brother). When requested by one other consumer 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 nicely. For example, a number of supposed responses from the chatbot simply look like transcripts taken from episodes of Democracy Now, the left-leaning information program hosted by Amy Goodman. When one consumer requested the chatbot a query about Jeffrey Epstein, the posted response included the next textual content:
“AMY GOODMAN: I wish 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 submit 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 submit, Watkins famous that he “wouldn’t take any medical or any recommendation from the AI for a very long time.” That’s definitely true and it is usually true that you simply most likely shouldn’t ask 8kun AI any questions until you wish to spend your days studying garbled nothings from the bowels of the web.