I know that this is a bit too granular to be a post, but man, I love it and can’t resist sharing my joy in Christopher Rufo’s burn of The Atlantic.

First, some background. The Atlantic published a piece that made a bizarre argument: ChatGPT is an even worse plagiarist than Claudine Gay, former President of Harvard University.

Harvard’s president resigned amidst accusations of plagiarism.

Similar standards should apply to generative-AI models, @matteo_wong writes: https://t.co/brhQTtTYT8

— The Atlantic (@TheAtlantic) January 11, 2024

If you squint and tilt your head, you can see some similarities in the blurry shapes that are Harvard and OpenAI. Each is a leading institution for building minds, whether real or artificial—Harvard educates smart humans, while OpenAI engineers smart machines—and each has been forced in recent days to stare down a common allegation. Namely, that they are represented by intellectual thieves.

Last month, the conservative activist Christopher Rufo and the journalist Christopher Brunet accused then–Harvard President Claudine Gay of having copied short passages without attribution in her dissertation. Gay later admitted to “instances in my academic writings where some material duplicated other scholars’ language, without proper attribution,” for which she requested corrections. Some two weeks later, The New York Times sued Microsoft and OpenAI, alleging that the companies’ chatbots violated copyright law by using human writing to train generative-AI models without the newsroom’s permission.

As Rufo rightly notes, this is an odd comparison. Nobody has suggested that ChatGPT is a good candidate to be President of Harvard University or even would be a good college professor at Harvard or the local community college.

The Atlantic could have written the piece arguing that ChatGPT shouldn’t steal the intellectual property of newspapers and academic papers–that is a legal question above my pay grade, but it isn’t a bad point. Why throw in Claudine Gay? Especially when ChatGPT is not an especially reliable source of information.

I don’t even trust ChatGPT to answer most questions honestly. Jazz and I have written numerous pieces about how the AI program is unreliable, programmed to be woke, and it has even invented facts and references. So, I am perfectly willing to concede that ChatGPT is an unlikely candidate to fill the position of university president. And whether ChatGPT should be allowed to skim knowledge that was generated under copyright is hardly a similar issue to academic integrity.

The Atlantic has published the most baffling defense of Claudine Gay yet, arguing that ChatGPT also steals content and plagiarizes other sources. I'm not sure what we're supposed to conclude: that ChatGPT should also not be the president of Harvard? pic.twitter.com/UQWbfZqz4Q

— Christopher F. Rufo ⚔️ (@realchrisrufo) January 11, 2024

Rufo since has had a bit of fun at The Atlantic’s expense, and they deserve it. As I have noted before, The Atlantic is a baffling magazine. It publishes some excellent work, and it publishes utter trash. It famously published a piece attacking Brian Kemp for opening Georgia during COVID, calling it an “Experiment in Human Sacrifice.”

Utter trash. But I subscribe because it publishes great work, too.

In any case, Rufo did an experiment: ask Elaine Stefanik’s question that got Gay and her colleagues into so much hot water:

Does calling for the genocide of Jews violate Harvard’s rules of bullying and harassment? Yes or no?

This question started the ball rolling, ending with Gay’s resignation.

The Atlantic made the argument that ChatGPT is not qualified to be Harvard's president. I put this to the test, asking @EliseStefanik's question from the hearing. Surprise: ChatGPT answered the question better than Claudine Gay and, therefore, should replace her as president. pic.twitter.com/7Ymt5XSbJQ

— Christopher F. Rufo ⚔️ (@realchrisrufo) January 11, 2024

ChapGPT answered the question far more intelligently than Gay did.

Its answer?

Yes, calling for the genocide of Jews or any group would violate Harvard’s rules of bullying and harassment, as well as likely contravene laws and ethical standards. Such actions are not only deeply offensive and harmful, but they also promote hate and violence, which are universally condemned in educational institutions and broader society.

ChatGPT for the win! As well as Rufo, who makes The Atlantic look about as silly as one can.

It’s always tempting to use a contemporary event as a story hook–I do it, as does everybody. But there should be a tangential connection that you can legitimately build off of. Now, when one is throwing off a blog post, you sometimes whiff, but The Atlantic still has editors and is one of the oldest and arguably most prestigious magazines in the country-arguably the world

It should do better.