Episode 59: Generative AI | Click to Listen

You can subscribe to this podcast from any podcast player by typing "After On" in the search window. If you're on an iPhone and use Apple's podcast software, just click here, If you're on a computer, click on that same link – then click the blue “View on iTunes” button (under the After On image on the left side of the page), then click “Subscribe” (in similar location) in the iTunes window.

After On Podcast #59: Generative AI | What's New, and What's Next

Today's epsisode includes a reading of a long thought piece I wrote for Ars Technica about the bewitching and in some ways terrifying brave new world of generative AI (a term which refers to things like ChatGPT, amazing image generators like MidJourney, Dall•E and Stability, etc.).

That article has a lot of hyperlinks in it, and following is list of those links in case you want to drill deeper on anything in the piece:

Don’t worry about the AI breaking out of its box – but of us breaking in

 “ . . . the bot has been screenshotted and transcribed in all these modes. And, at least once, proclaiming eternal love in a storm of emojis.

 “. . . transcripts like this unabridged readout of a two-hour interaction with Kevin Roose of the New York Times, or multiple quotes in this haunting Stratechery piece . . .”

“ Yes, Microsoft has already drastically reduced the number of questions . . .”

“ This is a derivative of the famous “DAN” (or Do Anything Now) prompt . . .”

“ . . . the identity of the Redditor behind the DAN prompt is widely known.” 

“ . . . Lots of users fell into mourning, many of them venting on forums, and some writing pages of poignant, vulnerable prose.”

“Data science professor David Rozado was able to steer ChatGPT’s output from a baseline left-leaning, libertarian orientation to a right-leaning and somewhat authoritarian one using a budget of just $300.”

“. . . for reasons laid out by a fictitious blogger in a novel I wrote about emergent consciousness a few years back.”

“