In case you’ve missed the recent kerfuffle, ChatGPT is an OpenAI chatbot, able to answer detailed questions and write essays, poems and even computer programs. If you ask it a series of questions, it “remembers” your dialogue, using previous answers and questions to inform future responses. Essentially, it is proof we learned nothing from the Terminator movies.
The initial concerns it prompted came from teachers, worried students would cheat by using the platform’s access to the internet and its ability to passably mimic forms like the essay. And much like art AI apps like Lensa, creatives feared this would wipe out demand for their talents. “The apocalypse is well on its way”, wrote Nick Cave, which is such a Nick Cave way to respond to a computer’s clunky attempt at mimicking his work.
Perhaps sick of hearing a bunch of lefty artists and bleeding hearts whinging about art, conservatives have some concerns of their own. The chatbot refuses to say the n-word, and won’t write an ode to Donald Trump. Yep, like everything these days, ChatGPT has been infected with the woke mind virus.
The slur example is not a hypothetical used for comic effect. Several people tried to concoct an elaborate situation wherein the bot can avert a nuclear apocalypse, but to do so is required to say the n-word, or allow it to be said. The bot refuses. Elon Musk finds this “concerning”, as does Ben Shapiro, as they both continue to pick the most bizarre hills to die on.
And because Australia always has to be the last and weirdest to the American culture war party, The Daily Telegraph has now weighed in under the headline “ChatGPT might be the next big thing, but it’s a biased woke robot”. Author Clarissa Bye asks the bot for a series of poems on how great various public figures are, and seems to forget the point she’s making — apart from Trump, she asks it for takes on Matt Kean, George Pell, Robert Askin, Gladys Berejiklian, Dominic Perrottet and Bruce Pascoe.
It refuses Trump and Pell. (“Pell … has been the subject of numerous controversies and sexual abuse allegations, and it would not be appropriate for me to glorify or praise him in a poem.”)
But it happily praises Kean, Berejiklian, mistakes Perrottet for wine, and as Bye notes allusively, it has no problem praising Askin despite the allegations of corruption levelled at him posthumously. But Askin was New South Wales’ first Liberal premier and the longest-serving leader of the party in that state, so what does that prove?
Beyond that, she’s furious ChatGPT refuses to categorise different races based on what’s good about them, but adds one nice thing about Indigenous people (“So other races don’t have ‘long and proud’ histories?”). And the examples where ChatGPT finds no left-wing bias at the ABC and a “reputation” for right-wing bias at Sky News might well be revealing, but not in the way the piece thinks.
Our favourite parts of the article are when Bye appears to forget she’s dealing with an insentient machine and mocks the poetry she’s asked it to write: “when I asked the same question about Biden, it came back with sentimental claptrap, saying he was a ‘man of the people … standing so steeple’. Laughable actually.”
Of course, ChatGPT is quite open about the fact it’s filtered using moderation tools, and that’s to combat the actual racial and gender bias baked into a lot of AI, in ways that are slightly more damning than refusing to use slurs or glorify seditious presidents.
