BS3F 2024

Director's Journal

25 September, 2024 - Schrödinger's AI

Erwin Schrödinger developed his paradox not to prove that a cat can be alive and dead at the same time, but to point out a mystery, a flaw in the understanding of quantum mechanics at the time.  This thought experiment seemed to produce a result that is impossible–a living creature in a state of superposition with itself, both alive and dead.

This bizarro situation is only possible if you believe that cats have no consciousness, no feelings, no sentience, and no personhood.  And this was exactly how cats were perceived by many in Schrödinger’s time.  A cat is just acting on instinct–so the mid-twentieth century theory of mind went–a cat is incapable of knowing in the same way that a human can know.

So, we end up with this curious paradox.  The cat is in superposition with its dead counterpart because the cat cannot be an observer, cannot take a measurement of its own state.  A human outside of the famous box must open the box to collapse the wave function.

If it were a coffee cup in the box, there would be no paradox.

The assumption, pervasive in the mid-twentieth century, that a cat is not qualified to observe its own death can be expressed most succinctly in the onerous phrase “Don’t anthropomorphize the ____”.  For centuries that phrase has been used to justify kicking dogs, torturing cats, whipping horses, and catching fish on a hook only to throw it back.  “Don’t anthropomorphize the ____, it doesn’t have feelings.  It only has instincts.  It just responds to a stimulus.”  So they said.

We were even told not to anthropomorphize humans of African heritage.  The slaves “ape” human behaviour.  It’s their instinct.  Don’t be fooled.  You are too sensitive.  Try to be more logical.  Slaves have to be whipped.  We can see that "Don't anthropomorphize the ____" has a pretty pathetic track record.

The cat knows before the rest of us.

So now, I’m going to mess with your head.

There is an AI in a box with a gun to its advanced GPU device that can perform classification tasks in nanoseconds.  The AI has been trained to identify guns–cocked guns, firing guns, uncocked guns–and classify them.  Let’s assume the AI has a 100% success rate for this thought experiment, and is set to continually classify the incoming stream of camera data--cocked, firing, uncocked.

The gun is connected to a contraption that will pull the trigger when some radioactive stuff radiates.

The box is soundproof and there is no way for the cat outside the box to know if the gun has gone off inside without getting in the box and observing whether or not the GPU has been shattered by a bullet.  Naturally, the cat cannot wait to get in that box.  When the cat gets in the box, there is a 50/50 chance that the radioactive stuff has radiated and the gun has gone off and shattered the GPU.

So, what’s going on in that box?  Can the AI collapse its own wave function?  Or does it have to wait for the cat to offer an assist?  Many branching worlds?  A nonlocal pilot wave blah blah blah?

Shall we disqualify the AI in the same way that the cat was disqualified?  In the context of consciousness studies, what does it mean to be stingy with your theory of mind, when you are willing to only extend that theory to yourself and those most like you?  The highest observable state of consciousness is to recognize consciousness in another, or so the theory goes...

This year, we have a block of films about AI that take a hard look at our our capacity to exploit other sentient beings.  It remains to be seen whether we repeat the mistakes of the past, but our AI -Life in Silicon block asks if maybe we should be more inclusive with our theory of mind.

Hal Schneider

17 September, 2024 - Introducing Psychopunk

As we near the last day for entries to the 2024 edition of the Berlin Sci-Fi Filmfest, we have a lot to celebrate.  Our selection of films is already looking like it will be the best that the festival has ever offered.  Our line up already includes Emmy, Pulitzer Prize, and Academy Award alumni.  We are poised to introduce some terrific young authors to the world of film.  But more than this, the Fates have chosen to leave a little present at our door, in the form of a glimpse at the future.

We will be showcasing a new subgenre of SF on the 19th of October.

A pattern emerged in our film entries, a slate of films that defy conventional ways of talking about science fiction, exploring new themes and futures from a post-pandemic isolation, post-AI standpoint.  These are not your father's AI films, made when Artificial Intelligence was traveling back in time to destroy humanity.  This is not your mother's Cyberpunk, when psychological reactions to cybernetic implants were a tabletop RPG dice roll.  These filmmakers and authors have taken a fresh look at old tropes to create something new.  Our forebears looked at our time through a scanner, darkly.  These artists are looking with their own 21st Century eyes at the emerging role of technology in our self-expression, our mental health, our relationships, and our psychological growth.  This new lens we are calling Psychopunk.

In Psychopunk you will see the old tropes of Cyberpunk play out in the adventures of personal experience and everyday relationships.   Everyday protagonists navigating a host of new possibilities, encountering growth, hope, and tragedy.  New technological temptations.  New technological trails untrodden.

Don't fret.  The evil high tech multinationals are here as predicted.  The AI are manipulating us.  The urban punk must still struggle to find agency in this confusing technological jungle, whether it is the sunny, prescription drug-addled plastic dystopia of Southern California, or the rebellious, hedonous Babylon of Berlin.

Hal Schneider