"Sophia", a film about a humanoid robot, won an award at #Mountainfilm. On stage, the one member of the production crew who went up to accept it (sorry, don't recall his role) cleverly used ChatGPT to write the acceptance speech and read it out loud on stage. It was a great gag, but where it went off the rails is when someone else on stage said it was scary in how accurate it was (paraphrasing).
But, no, it was not scary. In fact, it was the most blandly, generic, fill-in-the-blanks acceptance speech you can imagine โ which is to say, exactly what you expect from autocomplete-on-steroids. It was manifestly *not* any sort of vague example of "intelligence."
I got a look this weekend into how the mainstream world views "AI" and how effectively the people who most stand to benefit from this technology have been in framing the narrative.
If ChatGPT were really clever, it would've thanked Sheila Nevins.* (she had nothing to do with the film.) Now that would've been scary.
*Old doc filmmaker joke.
#AIHype