A week after The Guardian wrote a shortish piece breathlessly touting new writing AI, specifically a program called GPT-3, The Atlantic takes a deep dive in the shallow end of the pool.
At least Renée DiResta gives us unedited, albeit excerpted, material to read (the Guardian only had edited slices):
In addition to the potential for AI-generated false stories, there’s a simultaneously scary and exciting future where AI-generated false stories are the norm. The rise of the software engineer has given us the power to create new kinds of spaces: virtual reality and augmented reality are now possible, and the “Internet of things” is increasingly entering our homes. This past year, we’ve seen a new type of art: that which is created by algorithms and not humans. In this future, AI-generated content will continue to become more sophisticated, and it will be increasingly difficult to differentiate it from the content that is created by humans. One of the implications of the rise in AI-generated content is that the public will have to contend with the reality that it will be increasingly difficult to differentiate between generated content and human-generated content.
Really? That's supposed to impress me? DiResta says yes:
It’s somewhat disconcerting to have a machine plausibly imitating your writing style based on a few paragraphs—to see it mindlessly generating “thoughts” that you have had. Even though I knew GPT-3 was putting words together based on probabilities, I found myself giving it prompts when I was stuck on how best to express a concept or wondering how it would answer a question.
Delude me, is more like it. Or delude her, which is more disconcerting. I am reminded of the old fake computer psychiatrist ELIZA, somewhat smoother version, and nothing else, other than the number of people who deluded themselves about it, too.
As for the horrors of AI being used for propaganda writing? Well, if Russian trolls can be replaced with AI bots to "flood the zone" even more, or capitalist businesses in America doing the same to We the People, that is troublesome to a degree.
But nowhere near the breathlessness degree.
Also contra DiResta, the "new AI" (think The Who, don't get fooled again) still needs humans to "prompt" it, and the Guardian admitted its excerpts were edited.
And "tells" will still happen. Response time, or response sloppiness, on social media will be one of them. One human stringer can only run so many bots at a time.
As for letters to the editor? Ms. DiResta, astroturfing campaigns opened that barn door years if not decades ago, and better-staffed newspapers regularly screwed the pooch.
Moving beyond the media angle, though, which is somewhat what the Guardian does? It claims editing on its AI piece took less than a human piece. That, in turn, makes me wonder what level of dreck its writers, or freelance op-ed submitters, actually turn in.
Other than the narrow world of yet more media-industry job losses, when I look at this, am I worried? No. As for other AI claims the Guardian flouts, I've heard some of the classical music. B-grade Philip Glass at best. And, that includes the same minimalist framework. Nothing anywhere close to Mahler or Stravinsky.
No comments:
Post a Comment