OPINION:
Music video producer turned podcaster Robby Starbuck last week accused Google of defaming him with artificial intelligence. His lawsuit against the search engine giant is novel, but such complaints will likely become more common as Silicon Valley tycoons continue adding half-baked AI features into every product.
Tech titans are infatuated with AI, even though the technology conceals the reasoning and sources underlying its assertions. This makes it easy to manipulate, whether by accident or design. In this case, Hollywood lefties have held a grudge against Mr. Starbuck since he came out of the closet as a Republican.
According to Mr. Starbuck’s court filing, Google’s chatbot portrayed him as a “child rapist, a serial sexual abuser convicted of assault, one who engages in financial exploitation, one who engages in ‘black ops’ tactics such as illegal campaign finance practices, and a shooter — in short, as a monster.”
When asked about it, Google’s Gemini offered a robotic admission of error: “No, Robby Starbuck has not been found guilty of child rape and has no criminal record. The false accusations arose from defamatory statements generated by artificial intelligence systems owned by Meta and Google. … The AI systems ‘hallucinated’ and fabricated numerous defamatory statements, which have led to serious consequences for Starbuck and his family.”
Referencing the litigation, the search engine tacked on a handy disclaimer: “AI responses may include mistakes. For legal advice, consult a professional.”
AI is prone to getting things so horribly wrong because its models are trained on material poisoned by the left. The two most prominent ingredients for AI summaries of the “truth” are posts on the online message board Reddit and the digital encyclopedia Wikipedia.
Reddit is notorious for assigning left-wing moderators to silence inconvenient opinions. For instance, a United Airlines pilot who was tapped to help run an aviation discussion group with 300,000 members was recently caught celebrating Charlie Kirk’s assassination.
Vitriol-filled commentary likely infected AI’s take on Mr. Starbuck’s life. Wikipedia, the internet encyclopedia, fares no better because it relies on superusers to inject liberal propaganda into articles. Co-creator Larry Sanger denounced the site’s current management for abandoning Wikipedia’s founding principles of neutrality.
Mr. Sanger proposed a series of reforms he calls “nine theses” to restore fairness and balance. He says it’s time to stop determining what’s worth printing based on consensus. Real consensus is impossible when the site blacklists conservative news outlets, automatically excluding them from the conversation.
A panel of dominant but anonymous Wikipedia administrators passes final judgment on controversies. Because their identities are shielded, there is no accountability for misconduct. It also opens the possibility of special interest groups paying these gatekeepers under the table to influence outcomes.
Despite the flaws, Google has decreed Wikipedia worthy of appearing at the top of search results, magnifying the impact of its partisan editorial choices. Google is obviously fine with that, considering the company’s fawning support for Democratic presidential candidates.
Noting Mr. Sanger’s dissenting view, Elon Musk created Grokipedia as an alternative, “aiming for maximum truth-seeking over biased editing.” It debuted Monday. If the venture avoids partiality, it could become popular.
The public has grown weary of the left’s ceaseless narrative engineering campaigns. Mr. Starbuck is also tired, and he is demanding that Google pay him $15 million in compensation for the firm’s deranged assault on his reputation.
Hopefully, a combination of legal action and competition will encourage better behavior.
Please read our comment policy before commenting.