I wrote this article submission a couple weeks ago for submission in the Lucid Dreaming Experience magazine’s September 2023 edition (not yet released). The magazine website is lucidDreamingMagazine.com. I use the pen name Eleanor Cait for my LDE submissions.
Futurists suspect that artificial intelligence will, in the next few decades, become smarter than humans (“the singularity”). With the advancement of “smarter” artificial intelligence like ChatGPT, I would wonder if it’s possible with sufficiently advanced AI for a machine to dream. Some dream researchers describe the processing of dreaming as akin to defragging a computer’s hard drive, but even from what I remember in the late 1990s, earlier computers needed to defrag a lot less than humans need to dream.
There was a short-lived Nickelodeon cartoon (c. 2004) called “My Life As a Teenage Robot”. One episode involved said teenage robot, Jenny, having her programming updated to enable her to dream, after she envied her human friends’ ability to do that. That episode has stuck with me since then.
A much more well-known fictional robot is Data on Star Trek: The Next Generation. If he’s programmed to sleep and dream similarly to us, would he lucid dream, and would he always lucid dream when he sleeps? Considering Star Trek in all iterations takes place several centuries from now (as opposed to the aforementioned singularity, predicted to happen when a lot of today’s lucid dreamers are still alive), both humans and machines may have advanced to the point where lucid dreaming is used to promote exploration of our inner and outer worlds, and robots like Data would be programmed to lucid dream.
Although I don’t generally encounter modern technology within my lucid dreams right now, I would love to see what the likes of Jenny and Data are like as dream figures, and it is true that one of my own dream signs is a futuristic setting. As of yet, human lucid dreamers outside a sleep lab cannot access waking life smart tech functions, but when robots lucid dream, they likely could, be they controlled by a human or not. The same may apply to AI dream figures in our lucid dreams, just like how intelligent biological lucid dream figures seem to be able to access humanity’s collective consciousness (as described by Carl Jung).
If and when robots are programmed to lucid dream, I’d picture their lucid dreams to differ from ours in specific ways. This is because their neural programming was specifically created by humans (at least at first), unlike ours. Their lucid dreams may have significantly more mathematical formulae than I and other humans experience within our lucid dreams, such as binary code. I imagine they’d have fewer dream figures generally (both humans and computers) than we humans experience in our lucid dreams, due to robots’ significantly lower desire for companionship and socialization than humans generally need. Due to their programming, robots wouldn’t need to use reality tests to become and stay lucid - their lucid dream experiences would be akin to a human having a Wake Induced Lucid Dream (WILD), minus the initial bodily sleep paralysis period.
I feel that if robots were programmed how to lucid dream, they’d be able to simulate within a lucid dream the “taking over as the most intelligent terrestrial species” scenario so often seen in post-singularity science fiction, just like beginning lucid dreamers often take advantage of having superpowers once lucid. Will lucid dreaming robots be better or worse once the singularity happens? I figure we don’t need to wait until AI “gets that smart” – we can test these scenarios in our lucid dreams right now.
(I originally considered co-writing this article with ChatGPT, but I’ve recently come to realize my skepticism towards today’s AI being independently creative. Although it knows what creativity is, and knows how to write in different styles given an appropriate prompt, it cannot yet gain inspiration – much less from the lucid dreams it doesn’t have – from independent thought. Its speculations on its own future are purely based in what humans have put online so far. It learns solely from that broad database called the internet, and without humans contributing to that database, it could not learn, according to this lucid dreamer. However, I should mention that my techie fiancée disagrees with me on ChatGPT’s creative ability).