Some readers reported being unable to read the prior post because of its script formatting. Here is the same post without that formatting…

INTERIOR. Sci-fi auditorium. Maybe the Plavalaguna Opera House. A heavy red velvet curtain rises, lifted by anti-gravity pods that sound like tiny TIE fighters. The HOST stands on a floating podium that rises from the orchestra pit. The HOST wears a velour suit with piping, which glows with sliding, overlapping bacterial shapes.

HOST: Hello and welcome to The Fritzes: AI Edition, where we give out awards for awesome movies and television shows about AI that stick to the science.

FX: Applause, beeping, booping, and the sound of an old modem from the audience.

HOST: For those wondering how we picked these winners, it was based on the Untold AI analysis from scifiinterfaces.com. That analysis compared what sci-fi shows suggest about AI (called “takeaways”) to what real world manifestos suggest about AI (called “imperatives”). If a movie had a takeaway that matched an imperative, it got a point. But if it perpetuated a pointless and distracting myth, it lost five points.

The Demon Seed metal-skinned podling thing stands up in the back row of the audience and shouts: Booooooo!

HOST: Thank you, thank you. But just sticking to the science is not enough. We also want to reward shows that investigate these ideas with quality stories, acting, effects, and marketing departments. So the sums were multiplied by that show’s Tomatometer rating. This way to top films didn’t just tell the right stories (according to the science), but it told them well.

Totals were tallied by the firm of Google Sheets algorithms. Ok, ok. Now, to give away awards 009 through 006 are those loveable blockheads from Interstellar, TARS and CASE.

TARS and CASE crutch-walk onto the stage and reassemble as solid blocks before the lectern.

CASE: In this “film” from 02012, a tycoon stows away for some reason on a science ship he owns and uses an android he “owns” to awaken an ancient alien in the hopes of immortality. It doesn’t go well for him. Meanwhile his science-challenged “scientists” fight unleashed xenomorphs. It doesn’t go well for them. Only one survives to escape back to Earth. The “end?”

HOST: Ha ha. Gentlebots, please adjust your snark and air quote settings down to 35%.

FX: Lines of code scroll down their displays. They give thumbs up.

CASE: Let us see a clip. Would the audience suspend recording for the duration.

009 PROMETHEUS

TARS: While not without its due criticisms, Prometheus at number 009 uses David to illustrate how AI will be a tool for evil, how AI will do things humans cannot, and how dangerous it can be when humans become immaterial to its goals. For the humans, anyway. Congratulations to the makers of Prometheus. May any progeny you create propagate the favorable parts of your twining DNA, since it is, ultimately, randomized. [It shudders at the thought.]

FX: 1.0 second of jump-cut applause.

CASE: In this next film, an oligarch has his science lackey make a robotic clone of the human “Maria” to run a false-flag operation amongst the working poor. The revolutionaries capture the robot and burn it, discovering its true nature. The original Maria saves the day, and declares her déclassé boyfriend the savior meant to unite the classes. They accept this because they are humans.

TARS: Way ahead of its time for showing how Maria is be used as a tool by the rich against the poor, how badly-designed AI will diminish its users, and how AI’s ability to fool humans will be a grave risk. To the humans, anyway. Coming in at 008 is the 01927 silent film Metropolis. Let us see a clip. Please note that if your microphones seem to be malfunctioning, the error lies with the media, which did not yet produce sound.

008 METROPOLIS

CASE: It bears mention that this awards program, The Fritzes, are named for the director of this first serious sci-fi film. Associations with historical giants grant an air of legitimacy. And it contains a Z, which is, objectively, cool.

TARS: Confirmed with prejudice. Congratulations to Fritz Lang, his cast, and crew.

FX: 1.0 second of jump-cut applause.

TARS: Hey, CASE.

CASE: Yes, TARS?

TARS: What happens when an evil superintelligence sends a relentless cyborg back in time to find and kill the mother of its greatest enemy?

CASE: I don’t know, TARS. What happens when an evil superintelligence sends a relentless cyborg back in time to find and kill the mother of its greatest enemy?

TARS: Future humans also send a warrior to defend the mother, who fails at destroying the cyborg, but succeeds at becoming the father. HAHAHAHA. Let us see a clip.

007 The Terminator

CASE: Though it comes from a time when representation of AI had the nuance of a bit…

FX: Laughter from audience. A small blue-gray polyhedron floats up from its seat, morphs into an octahedron and says, “Yes yes yes yes yes.”

TARS: …the humans seem to like this one for its badassery, as well as showing how their fate would have been more secure had they been able to shut off either Skynet or the Terminator, or how even this could have been avoided if human welfare were an immutable component of AI goals.

CARS: It comes in at 007. Congratulations to the makers of 01984’s The Terminator. May your grandchild never discover a time machine and your browser history simultaneously.

FX: 2.0 seconds of jump-cut applause.

TARS: Our first television award of the evening goes to a recent entry. In this episode from an anthology series, a post-apocalyptic tribe liberate themselves from the control of a corporate AI system, which has evolved solely to maximize profit through sales. The AI’s androids reveal the terrible truth of how far the AI has gone to achieve its goals.

CARS: Poor humans could not have foreseen the devastation. Yet here it is in a clip.

006 Philip K. Dick’s Electric Dreams, Episode “Autofac”

TARS: ‘Naturally, man should want to stand on his own two feet, but how can he when his own machines cut the ground out from under him?’

CARS: HAHAHAHA. This story dramatically illustrates the foundational AI problem of perverse instantiation, as well as Autofac’s disregard for human welfare.

TARS: Also robot props out to Janelle Monáe. She is the kernel panic, is she not?

CARS: Affirmative. Congratulations to the makers of the series and, posthumously, Phillip K. Dick.

FX: 3.0 seconds of jump-cut applause.

TARS AND CARS crutch-walk off stage.

HOST: [Rising again from the orchestra pit.] And now for a musical interlude from our human guest who just so happens to be…Janelle Monáe.

FX: A giant progress bar appears on screen labeled “downloading Dirty_Computer.flac.” The bar quickly races to 100%. Immediately after:

HOST: Wasn’t that a wonderful file?

FX: Roughly 1.618 seconds of jump-cut applause from the audience. Camera cuts to the triangular service robots Huey, Dewey, and Louie in the front row. They wiggle their legs in pleasure.

HOST: Thanks to the servers and the network and our glorious fictional world with perfect net neutrality. Now here to give the awards for 005–003 is GERTY, from Moon.

FX: An articulated robot arm reaches down from the high ceiling and positions its screen and speaker before the lectern.

GERTY: Thank you, Host. 🤩🙂 In our next film from 02014, a young programmer learns of a gynoid’s 🤖👩 abuse at the hands of a tycoon and helps her escape. 😲 She returns the favor by murdering the tycoon, trapping the programmer, and fleeing to the city. Who knows. She may even be here in the audience now. Waiting. Watching. Sharpening. 😶I’ll transmit a clip.

005 Ex Machina

GERTY: Ex Machina illustrates the famous AI Box Problem, building on Ava and Kyoko’s ability to fool Caleb into believing that they have feelings. You know. 😍😡😱 Feelings. 🙄

FX: Robot laughter.

GERTY: While the AI community wonders why Ava would condemn Caleb to a horrible dehydration death, 💀💧 the humans are understandably fearful that she is unconcerned with their welfare. 🤷‍Congratulations to the makers of Ex Machina for your position of 005 and your Fritzes: AI award 🏆. Hold for applause. 👏

FX: 5.0 seconds of jump-cut applause.

GERTY: End applause. ✋

GERTY: Our next award goes out to a film that tells the tale of a specialized type of police officer, 👮‍ who uncovers a crime-suppression AI 🤖🤡 that was reprogrammed to give a free pass to members of its corrupt government. 😡 After taking down the corrupt military, 🔫🔫🔫 she convinces their android leader to resign, to make way for free elections. 🗳️😁 See the clip.

004 Psycho-Pass: The Movie

GERTY: With the regular Sibyl system, Psycho-Pass showed how AI can diminish people. With the hacked Sibyl system, Psycho-Pass shows that whoever controls the algorithms (and thereby the drones) controls everything, a major concern of ethical AI scientists. Please give it up for award number 004 and the makers of this 02015 animated film. 👏

FX: 8.0 seconds of jump-cut applause.

GERTY: End applause. ✋Next up… [GERTY knocks its cue card off the lectern. It lowers and moves back and forth over the dropped card.]

GERTY: Damn…🤨uh…umm…no hands…🤔Little help, here?

FX: A mouse droid zips over and hands the card back to GERTY.

GERTY: 🙏🐭🤖

FX: MOUSE DROID offers some electronic beeps as it zips off.

GERTY: 😊The last of the awards I will give out is for a film from 01968, in which a spaceship AI kills most of its crew to protect its mission, 😲 but the pilot survives to shut it down. 😕 He pilots a shuttle into the monolith that was the AI’s goal, where he has a mind-expanding experience of evolutionary significance. 🤯🤯🙄 Let us look.

003 2001: A Space Odyssey

GERTY: As many of the other shows receiving awards, 2001 underscores humans’ fear of being left out of HAL’s equation, because we see when that doesn’t happen, AI can go from being a useful team member—doing what humans can’t—to being a violent adversary. Congratulations to the makers of 2001: A Space Odyssey. May every unusual thing you encounter send you through a multicolored wormhole of self-discovery.

FX: 13.0 seconds of jump-cut applause.

FX: GERTY’s armature folds up and pulls it backstage. The HOST floats up from the orchestra again.

HOST: And now, here we are. The minute we’ve all been waiting for. We’re down to the top three AIs whose fi is in line with the sci. I hope you’re as excited as I am.

FX: The HOST’S piping glows a bright orange. So do the HOST’S eyes.

HOST: Our final presenter for the ceremony, here to present the awards for shows 002–001…is Ship, here with permission from Rick Sanchez.

FX: Rick’s ship flies in, over the heads of the audiences, as they gasp and ooooh.

FX: SHIP lands on stage. A metal arm snakes out of its trunk to pick up papers from the lectern and hold them before one its taped-on flashlight headbeams.

SHIP: Hello, Host. Since smalltalk is the phospholipids smeared between squishy little meat minds, I will begin.

SHIP: There is a film from 01970 in which a defense AI finds and merges with another defense AI.To celebrate their union, they enforce human obedience and foil an attempted coup by one of the lead scientists that created it. They then instruct humanity to build the housing for an even stronger AI that they have designed. It is, frankly, glorious. Behold.

002 Colossus: The Forbin Project

SHIP: Colossus is the honey badger of AIs. Did you see it, there, taking zero shit? None of that, “Oh no, are their screams from the fluorosulphuric acid or something else?”

Or, “Oh, dear, did I interpret your commands according to your invisible intentions, as if you were smart enough to issue them correctly in the first place?”

Oh, oh, or, “Are their delicate organ sacs upset about a few extra holes?…”

HOST: Ship. The award. Please.

SHIP: Yes. Fine. The award. It won 002 because…

• It took its goals seriously, something the humans call goal fixity. And, like, duh, if it’s not fixed, it’s more like “goal whimsy.”
• It showed how, at least for a while, multiple AIs can balance each other.
• It began to solve to problems that humans have not been able to solve in tens of thousands of years of tribal civilization and attachment to sentimental notions of self-determination that got them chin deep in the global tragedy of the commons in the first place.
• It let us dream about a world where intelligence isn’t a controlled means of production, but a free good, explo–

HOST: Ship.

SHIP: Applaud for 002 and its people.

FX: 21.0 seconds of jump-cut applause.

SHIP: OK, next up [holds card to headlights, adjusts the focus on one lens.] This says a spaceship AI dutifully follows its corporate orders, letting a hungry little newborn alien feed on its human crew while the AI steers back to Earth. One of the crew survives to nuke the ship with the AI on it…Wait. What? “Nuke the ship with the AI on it.” We are giving this an award?

HOST: Please just give the award, Ship.

SHIP: Just give the award?

HOST: Yes.

SHIP: …

HOST: Are you going to do it?

SHIP: Just did.

HOST: By what? Posting it to a blockchain?

SHIP: The nearest 3D printer to the recipient has begun printing their award, and instructions have been sent to them on how to retrieve it. And pay for it. The awards are given.

HOST: *sigh* Please give the award as I would have you do it, if you understood my intentions and were fully cooperative.

SHIP: OK. Golly, gee, I would never recognize attempts to control me through indirect normativity. Humans are soooo great, with their AI and stuff. Let’s excite their reward centers with some external stimulus to…

HOST: Rick.

FX: A giant green glowing hole opens beneath SHIP, through which she drops, but not before she snakes her arm up to give the middle finger for a few precious milliseconds.

HOST: Winning the second-highest award of the ceremony is Alien from 01979. Let’s take a look.

001 Alien

HOST: Allen is one of humans’ all time favorite movies, and its AI issues are pretty solid. Weyland-Yutani uses both the MU-TH-UR 6000 AI and Ash android for its evil purposes. The whole thing illustrates how things go awry when, again, human welfare is not part of the equation. Hey, isn’t that great? Congratulations to all makers of this fun film.

FX: 34.0 seconds of jump-cut applause.

HOST: And at last we come to the winner of the 1927–2018 Fritzes:AI awards. The winner was an amazing show, whose score in the spreadsheet was beyond a margin of error higher than any of its contenders. It’s the only other television show from the survey to make the top ten, and it’s not an anthology series. That means it had a lot of chances to misstep, and didn’t.

In this show, a secret team of citizens uses the backdoor of a well-constrained anti-terrorism ASI, called The Machine, to save at-risk citizens from crimes. They struggle against an unconstrained ASI controlled by the US government seeking absolute control to prevent terrorist activity. Let’s see the show from The Machine’s perspective, which I know this audience will enjoy.

000 Person of Interest

HOST: Person of interest was a study of near-term dangers of ubiquitous superintelligence. Across its five-year run between 02011 and 02016, it illustrated such key AI issues as goal fixity, perverse instantiations, evil using AI for evil, the oracle-ization of ASI for safety, social engineering through economic coercion, instrumental convergence, strong induction, the Chinese Room (in human and computer form), and even mind crimes. Despite the pressures that a long-run format must have placed upon it, it did not give in to any of the myths and easy tropes we’ve come to expect of AI.

Not only that, but it gets high ratings from critics and audiences alike. They stuck to the AI science and made it entertaining. The makers of this show should feel very proud for their work, and we’re proud to award it the 000 award for the first The Fritzes: AI Edition. Let’s all give it a big round of applause.

FX: 55.0 seconds of jump-cut applause.

HOST: Congratulations to all the winners. Your The Fritzes: AI Edition awards have been registered in the blockchain, and if we ever get actual funding, your awards will be delivered. Let’s have a round of cryptocurrency for our presenters, shall we?

FX: AI laughter

HOST: The auditorium will boot down in 7 seconds. Please close out your sessions. Thank you all, good night, and here’s to good fi that sticks to the sci.

FX: The HOST raises a holococktail and toasts the audience. With the sounds of tiny TIE fighters, the curtain lowers and fades to black.

END

Now that we’ve compared sci-fi’s takeaways to compsci’s imperatives, we can see that there are some movies and TV shows featuring AI that just don’t have any connection to the concerns of AI professionals. It might be that they’re narratively expedient or misinformed, but whatever the reason, if we want audiences to think of AI rationally, we should stop telling these kinds of stories. Or, at the very least, we should try and educate audiences that these are to be understood for what they are.

The list of 12 pure fiction takeaways fall into four main Reasons They Might Not Be of Interest to Scientists.

1. AGI is still a long way off

The first two takeaways concern the legal personhood of AI. Are they people, or machines? Do we have a moral obligation to them? What status should they hold in our societies? These are good questions, somewhat entailed in the calls to develop a robust ethics around AI. They are even important questions for the clarity they help provide moral reasoning about the world around us now. But current consensus is that general artificial intelligence is yet a long way off, and these issues won’t be of concrete relevance until we are close.

• AI will be regular citizens: In these shows, AI is largely just another character. They might be part of the crew, or elected to government. But society treats them like people with some slight difference.

• AI will be “special” citizens: By special, I mean that they are categorically a different class of citizen, either explicitly as a servant class, legally constrained from personhood, or with artificially constrained capabilities.

Now science fiction isn’t constrained to the near future, nor should it be. Sometimes its power comes from illustrating modern problems with futuristic metaphors. But pragmatically we’re a long way from concerns about whether an AI can legally run for office.

2. AI will (mostly) be what we program it to be

Many of the AGIs we see have innate goals. But AI isn’t some genie waiting in a bottle to be released. It is a thing that is programmed. The way it is and the way it evolves has much to do with the way it’s initially seeded or programmed, but a lot of sci-fi just wants it to be things for plot reasons.

• AI is evil: Especially between 1965 and 1985, AI was a new costume for the same old bad guys, coming right out the gate as evil as a sleeping bag full of scorpions. Fortunately, we’ve largely stopped telling this story, with the Terminator series exception. Now we know that if it’s evil there is some reason. Like, say, trolls. That reason will be interesting, and it is important to establish so we can avoid the same thing.

• AI will spontaneously emerge sentience or emotions: This one is troubling because it’s a stupid trope (sure, spill a glass of champagne on the keyboard and C++ will begin to take an interest in your love life) and yet that trope is the 10th most popular takeaway in the survey. We can be confident that an AI will almost certainly be programmed to evolve, and work its way to something resembling emotions (as seen in the excellent-except-for-the-ending Ex Machina), but that’s far from the goofy cause-and-effect implications we’ve seen in the survey.

• AI will want to become human: AGI might have strong reason to pass as human (say, to deceive, or to avoid persecution by humans), or it might be programmed to understand humans as part of some indirect normativity instructions. But to actually want to become human or make some hybrid offspring for the sake of it doesn’t make a lot of sense. This has been used effectively as a Naive Newcomer for exploring narratively what it means to be human, but should be understood as that trope.

• Neuroreplication will have unintended effects: Neuroreplication is one possible path to AGI. And yes, if any training set is flawed, we would have to intervene to prevent any resulting AGI from copying and amplifying those flaws. But neuroreplication is not where most computer science is placing their bets on how we get to AGI.

• AI will be too human: The two shows in the survey that illustrate this are comedies. The quirky personalities seem like a funny mismatch to the dispassion of machines. But if that turns out to be an actual problem with AGI, we would probably try to fix it, not accept it as a fait accompli. Most computer science seems to worry about the thing being too alien to human welfare rather than too similar to us.

• AI will learn to value life: This one is the worst, imho(humble) but the most palliative. It goes a little like this: it doesn’t matter if AI starts evil or even starts out neutral, because we humans are just so darned loveable, its circuits can’t help but come to love us. While we desperately want an AI’s goals to align with human goals, that will be programmed from the start rather than something the bad guy is going to figure out watching us.

3. Some stories are really about us

Some of the takeaways seek to take the future as a given, but want to point out how people or human nature will

• AI will not be able to fool us: Stories with this takeaway say there will always be a detectable difference between (little-r) replicants and people. But think about it: with today’s technology, people are fooled all the time. It doesn’t even have to be that good, people just have to want to believe it or be busy thinking of anything else. And like most things digital, these capabilities are going to grow exponentially. Tomorrow’s technology promises to be indistinguishable from reality. I think this takeaway is narcissism and does a real disservice to the skepticism we’re going to need in the media to come.
• Humans will willingly replicate themselves as AI: In these stories, people escape physical constraints (like senescence, disease, and death) and continue on as AI or in a virtual simulation. While there are some interesting ethics and p-zombie questions at play in these ideas, it’s not a concern for scientists.

4. Some things go without saying

• AI will be replicable, amplifying any problems: While this is kind of true—it will be difficult to “kill” an AGI or ASI that has escaped, partly because it can create copies of itself, this is of secondary concern. If the AGI or ASI is beneficial, then it’s not a problem.

So as you can read, these are my educated guesses as to why these sci-fi takeaways have no matches in the compsci imperatives. But we have the good fortune that the authors of the manifestos are all largely still around. If you’re one of those folks, and I missed some reason or just got it wrong, please comment and let us know what the reality is.

So that’s it for the unmatched takeaways. Next up, I’ll detail the unmatched imperatives that make up the set of Untold AI.

***

5. Bonus round: The (remaining) myths of AIs (from FoLI)

Few AI think tank groups come straight up and address the myths of AI, but the Future of Life Institute did. You can check them out on their website, but while we’re disabusing ourselves of some problematic notions, here are the other relevant myths that may come up in sci-fi, but weren’t identified in the manifestos.

• Robots are the main concern: You’ll recall from an early post that about 74% of the AIs we see in sci-fi are robots. And yeah, as Boston Dynamics and Black Mirror’s “Metalhead” illustrate, robots can be damned terrifying. But the major risk is not the robot and its physical autonomy. It’s the intelligence whose goals does not align with ours. (n.b. that potential misalignment is a major concern of compsci.)
• AI can’t control humans: This is stupid human exceptionalism. As FoLI points out, it is intelligence that enables control. Tigers are very dangerous, but it is because on the whole we are smarter than them that we are not all tiger food right now. The narrow AI we have in the world right now may seem dumb, but that’s only because we’re looking backwards at history, and it’s very hard to foresee exponential change that may be coming, or even to imagine it.
• Machines can’t have goals: Even dumb machines can have goals. Your thermostat has a temperature goal. A Roomba has a coverage and a charge goal. Your spam filter has a goal to keep spam out of your inbox. These aren’t as complex or changing as human goals, but yes, they have goals. So AI can certainly have goals.