Police light

6 May 2013 by Christopher Noessel

This post (the first in what is going to amount to The Fifth Element Police Week. What is this, sweeps?) is going to veer to the edge of interaction design, getting into the Venn overlap of industrial design and wearable tech.

The police seen throughout New York each wear uniforms that feature a large, circular, glowing light over the right side of their chest.

There are only two things to say that’s positive about this police light. One: Yes, it looks cool. Two: It certainly gives narby citizens a clear, attention-getting signal that something is up. This might be OK for community relations officers, who are only ever interfacing with the public. But when it comes to dealing with actual criminals, it’s a terrible idea.

It’s a terrible idea because of its placement

Imagine this scene from the chief’s perspective. When he addresses Leeloo down the pipe as she’s standing on the ledge of the building, he is in an isosceles stance, with his shoulders perpendicular to the target and his weapon held in front of his heart. This common stance would place the weapon directly in the glow of the circle. This means that his forearms and weapon will have the brightest illumination in his field of vision and be distracting. This might be manageable by coating his uniform and the back of the weapon with a super black coating to absorb much of this light. But, depending on the distance of the target, it is also likely to place the perp in shadow, making them harder to see and harder to hit.

Looking at the officer on the right, we see he is taking a different stance. He is “bladed” to the target, closer to a Weaver stance, with his body turned a bit sideways. This stance turns the light to the adjacent wall, which minimizes the backscatter and perp-shadow effects, but also aims the light toward his fellow officer, possibly distracting him or her. That’s a pretty crappy design. But wait, it gets worse.

It’s a terrible idea because it’s a giant, glowing target

What’s worse is to imagine the scene from the perspective of the perp, say, the Mondosahwans in the airport. They want to specifically shoot the police in the crowd and all they have to do is shoot towards the glowing discs. That’s right, the police in 2263 are actually wearing attention-drawing targets. Admittedly, if you are going to get shot in the line of duty, you’d rather draw fire away from the head to a place with a solid slab of bone and lots of body armor. But why draw their fire in the first place?

As we saw in another post, Zorg believes in the fallacy/parable of the broken window, and so favors a bit of destruction that encourages market activity. We also know from the film that he has a lot of control over the NYPD. It might be that he’s deliberately sabotaging the police through this design to encourage the sale of more body armor and weapons, but are we to believe that the cops themselves are willing to go along with this? C’mon. They’re smarter than this.

Improve it with a little bit of smarts

Outfit the light with a little agentive smarts, and most of these problems could be fixed. The light could simply dim when it’s counterproductive to have it illuminated. Proximity sensors can sense when the officer’s arm is in the way. Context aware sensors can sense when it might blind another officer. It would take a lot of smarts to know when the officer is being targeted by a weapon, but certainly simple audio sensors should shut it off in the sound of gunfire.

5E-opedia: Search

3 May 2013 by Christopher Noessel

Leeloo learns about the facts of the human race which she is destined to save through an online encyclopedia available to her in many places: in Cornelius’ home, the spaceship to Fhloston Paradise, and aboard Zorg’s ship. Three modes are seen for it. Today we discuss the third mode, which is to search for an in-depth topic.

Search

When Leeloo experiences full-scale combat with Zorg and the Mangalores aboard Fhloston Paradise, she grows curious about war. On the route back to Earth aboard Zorg’s ship, she once again returns to the online encyclopedia she’s been referencing throughout the film. When she sits down, it just so happens that the system is in the middle of the W topics. It is amid “we*” and “wh*” words. “Weapon” is at the top, so maybe that’s what Zorg was looking for.

To access a particular topic not on screen, she simply begins typing. She types “WAR,” the letters filling the screen in green all-caps, and the entry for war begins playing. This entry is different than the prior one seen on martial arts. This is simply a series of still images presented serially, around four dozen that culminate in an image of the French test of an atomic weapon at Mururoa Atoll.

Two small nuances to note. The first is that we don’t see a result of possible search results. Like Wikipedia, there is a main entry for war, and it presumes that’s the one she means. If it’s wrong, she can interrupt. That’s a smart default that will work in most cases.

The second is that we don’t see or hear Leeloo hit an “enter” key after she finishes typing “war.” (The other keys each emit a small beep.) How did the system know she wasn’t continuing on to “warrior” or “warship”? A smart system would be able to interpret the pause after the “r” as a likely end, once it passes an outer threshold for her typical typing speed, and begin to show her the “war” entry. Then, if she continued to add another letter just outside that threshold, it could evaluate the string. If it might be a continuation, like typing an “s” for “warship” it could pause the display and wait. If a continuation wouldn’t make any sense, like “warx,” it could presume she was entering a new word beginning with “x” or help her recover in case it was just a plain old typo.

Interestingly, this is kind of the way Google Instant search works. Did the designers for The Fifth Element accidentally invent it 13 years ahead of Google?

Despite that cool possibility, I have to ding this entry for not really explaining anything. Some aren’t really about war but about terror, such as the image of the burning cross at a KKK rally. But even for the others, yes, they are horrific images. And they are a stinging reminder of the horrors that accompany war. But they really only work for someone with the prior knowledge of what they describe. Steve McCurry‘s haunting image of a tank in Kuwait, for instance, inspires despair only if you know the full background story of that war, and this sequence certainly does not provide it to Leeloo.

Ultimately, regardless of the mode this encyclopedia is in, it is a cinematic conceit that we should not take as a good example of rapid learning for the real world.

5E-opedia: In-depth topic

2 May 2013 by Christopher Noessel

In-Depth Topic

Leeloo can understand each item in the topic lists as they fly past. If she sees a topic that interests her in particular, she can press a button to find out more about that topic in more detail. (We don’t see the button, we just hear it.) Given that she’s looking at a screen of at most 66 and at least 4 options, and we don’t see a selection indicator, it’s anyone’s guess as to how she does this. Later we’ll see that she has a QWERTY keyboard to search for a particular word, and we don’t see that same search interface here, so it’s something other than that.

Once she indicates that she’s interested in martial arts, the entry fills the screen. The screen is a mix of a paragraph of text, images zooming around, and subtopics writ in large red majuscule letters scrolling past: KRAV CONTACT, SUMO, WRESTLING, SAVATE, KUNG FU, JU JITSU, NINJITSU, WRANG DO, FULL CONTACT… A still image of Bruce Lee from Enter the Dragon appears. This style of still-image and animated-text continues to play in a watch-and-learn way until it’s done, and then returns to the topic list.

Here, as before, I am examining things unmeant for examination. Still, I have a job to do. In the diegesis of the film, the text flies by too quick for anyone but a perfect Mondoshawan to read. But here in the real world, I hit pause. There I learned that the paragraph of text in the background has nothing to do with martial arts. We only see snippets, but they read as follows. (Please post your short sci-fi stories that can make sense of these lines in exactly this same order.)

: a hindu thus
talks to hi[s] troops about taking
d takes on a persona of its own.
monster, if it wants to live, have
loved. We then get a news flash
cult (think Waco siege coverage)

This little bit of text reads much more like a script than an encyclopedia entry. Like it was a bit of text just lying around on someone’s computer. In any case it would not help Leeloo learn Jeet Kun Do in the slightest.

On the right side of the screen (see above) we also see a vertical green rectangle. At the top is the number 5, bookended with arrows. Below that is a graph, a set of thumbnail images (whose captions are too small to read) are linked by right-angle connecting lines, like what you might see in a tech-tree for a real-time strategy video game.

When the display shifts to showing the subtopics, this green area changes. The 5 changes to a dot, and a grid of circular icons appears, each with a green rectangle to its right. The left column of icons is hard to decipher, but the right column of icons looks like control buttons one might expect: More detail, next in sequence, prior in sequence, zoom out, zoom all the way out, fast forward. Missing are common controls for video such as pause and play. A the bottom is a button labeled “EDIT”. This control panel is not seen in use.

It’s still about the learning, stupid

That stuff on the left is pointless. Of course that bit from a script is goofy. The animated stuff might be interesting for getting someone kind of excited about the topic, or maybe to remember how awesome martial arts (that they already knew about) are, but for learning any of it from a computer screen, she would have been better off spending time on youtube. Even the subtopics make no sense. Sure, they’re all martial arts, but what’s the order? Not alphabetical. Not age. Savate (18th century) is between wrestling and Sumo, both far more ancient. It’s not even a list of the same scope of thing. Aren’t Krav and Full Contact different translations for the same thing? Anyway, learning the vocabulary of a domain is only a rudimentary first step to actually learning it, much less performing it. Good thing she’s “perfect.”

The first green area on the right does actually seem useful for learning. It’s an abstract representation of how some things fit together. There’s a relationship implied between parts. It may also provide a map to a bigger picture in which this particular topic fits. That’s actually pretty useful and even Wikipedia adopts it for entries that fit into larger domains of knowledge. So, OK, we’ll cut it some slack there.

The second green area, even though I’m doing a lot of inference there from icons, also seems like it might be pretty useful. It’s too bad we don’t get to see it in action.

Better for Leeloo’s purposes of learning a topic—even if you did it blazingly fast—would be to provide her a definition, a bit about the history, and then some blazingly fast how-tos of modern practice augmented with the principles at work in each of the examples.

5E-opedia: Watch and learn

1 May 2013 by Christopher Noessel

Watch-and-learn

When we first see her using the (unnamed) encyclopedia, she is simply watching four columns of words quickly scroll by. The words are arranged alphabetically, top-to-bottom before continuing to the next column. There is a large, blinking, lower-case letter reversed out of a white square in the lower left. Near the middle of the screen, a thick bracket emphasizes four of the words from the screen and red lines connects each of the words to a large image on the right side of the screen. The words and pictures fly by at a rate that’s impossible for us mere humans to follow, but Cornelius assures David that she”s ‘learning our history…the last 5,000 years she’s missed. She’s been asleep for quite a while, you know.”

This mode is all passive. When in the scene Leeloo goes to grab a turkey from the kitchen to bring back and eat, the screen is still moving with no one in front of it. We see a menu of capital letters, and a selection moving from “A” to “M”, by itself.

The Tyranny of Pause

Here I’m going to have to break the usual stance with which I review interfaces. That is, I usually treat each interface as if it was perfectly as it should be in the world of the movie or TV show, trying to willfully ignore its speculative nature. (That’s how we can make it relevant to our real world work.) But here, there’s just too much that’s broken in the content to make any sense. You only notice it when you slow the movie down to read and examine the screens, so this is totally unfair. But then again, the DVD format had been in the world for two years by the time The Fifth Element came out, so there’s not a great excuse for playback technology. They could presume it would eventually be paused.

First off, the words in the lists are repeated. The first column is identical to the third. The second is identical to the fourth. I can’t imagine a good reason why this would help a reader. I was hoping maybe there was some autostereogram thing going on, but no. It just adds noise.

Secondly, the vast majority of images have little to do with the words to which they’re connected. “Me” points to a halved cantaloupe filled with blackberries. (See the image above.) “Maunder” points to an image of a woman’s softly parted, lipsticked lips. What on earth is Leeloo meant to learn from that? (Also I think it’s high time we bring back maunder into common usage.)

Worse, the images repeat. So the same picture of a chimpanzee connects to both “meadow” and “matriarch.” The same picture of a nose (flipped horizontally once) connects to “nav(vv?)y,” “nefarious,” and “negate.”

Also the same word may appear multiple times, connected to different pictures. “Maw” points once to a mouth, which is sensible, but once to a full-body portrait of a model in a little black dress. I’m all for polysemy and homonymy, but this just makes no sense.

Only once does the connection make absolute sense, as “Napoleon” points to François Gérard’s 1804 portrait of “Napoleon Bonaparte, Emperor of France, at Malmaison.”

Now I understand the tough challenge the interface designers faced: They probably had zero time, the damned thing had to look…well…encyclopedic, it had to make Leeloo look like a learning machine, no one had the budget or time to create new media for all these entries. Plus Larry Lessig wouldn’t found the Creative Commons organization for four years later. But…still.

Sometimes the words are arbitrarily cut off, as in “mayonnais” and “masturbat.” (Both the prurient-minded and those prone to conspiracies about subliminal influence may note that the word “masturbat~” appears three times, twice as a full word, and once cut off in this way.)

So, in short, unless she’s studying the Dada Encyclopedia, this display just makes no sense.

Um…learning?

Even if the images didn’t repeat, the words didn’t repeat, the images made sense for the words, and words appeared fully spelled out, it’s a ridiculous display for what Leeloo is trying to do. At best, this might be able to teach her the written words for some concrete things. But even this is doubtful. How does she know that “Napoleon” as a word means a particular individual rather than the word for a painted portrait, or the name of the uniform being depicted? Without going too far into history or learning theory, we can say that Leeloo would need exposure to some propositional language to understand history as interrelated events occurring across time, and an alphabetical list of words and pictures just isn’t enough.

ZF-1

30 Apr 2013 by Christopher Noessel

Amongst its many holdings (including taxi cab companies) Zorg industries manufactures weapons, including their flagship weapon, the ZF-1. It has a great many features. It stores as a sealed pod, and can be activated by a remote control. With a press of a button, shielding retracts and parts extend so it can be handled like a traditional small arms weapon.

Zorg makes a pitch to the Mangalores for the ZF-1, so we’ll just let his own words sell it.

It’s light. The handle is adjustable for easy carrying. Good for righties and lefties. Breaks down into four parts. Undetectable by x-ray. Ideal for quick, discreet interventions. A word on firepower: Titanium recharger. 3,000 round clip with liquid bursts of 3-to-300.”

Next he pitches something quite unique to the weapon.

“With the Replay™ button—another Zorg invention—it’s even easier. One shot…and Replay™ sends every following shot to the same location…”

As he turns and points the weapons at the Mangalores, the ammunition arcs around to home in on the first shot.

But wait, don’t answer. The ZF-1 has other features as well.

…And to finish the job, all the Zorg oldies but goldies: Rocket launcher, arrow launcher with explosive, poisonous gas heads (very practical), our famous net launcher, the always-efficient flamethrower (my favorite), and for the grand finale, the all-new Ice Cube™ System.

After the Magalores fail to uphold their end of the bargain, Zorg leaves them to play with the weapons. As they do, one discovers that the glowing red button on the side is actually an explosive self-destruct.

Analysis

I know Mangaglores are not meant to be shining examples of intellect, but if I was considering a purchase, I would yes, compliment the incredibly nifty technology of Replay, but follow it up with four more important questions about the design of the thing.

First, Mr. Zorg, what good is the remote control? Doesn’t this make the weapon hackable remotely? Isn’t that device easy to misplace? What on-weapon means do we have to unlock it?

Second, how are you selecting from among the six different types of ammunition?

On the exterior, we only see that red button. There might be some other subtle switches somewhere on the exterior, but you had to support the weight of the device with your left hand, so it’s fairly immobilized and I didn’t see you moving it. Unless it can only fire in exactly the order we saw, there’s got to be some other control. With your right hand hidden up inside the weapon, there must be other activation switches there. What switches are tucked up in there that are easy to differentiate by touch and easy to activate with your palm remaining against the grip?

Third, there’s that red button. Sure, who wouldn’t want to carry around a device that could erupt as an all-consuming fireball, but I notice that it doesn’t have a safety cover on it, gives no pause or warning during which the command can be retracted, and draws attention to itself by its glow. Isn’t that going to be increadibly easy to, you know, accidentally kill all my troops?

Fourth, during the demonstration we got a good glimpse at the front of the weapon. It’s got animated, blinking red LEDs whose pattern merges together to form a bright red diamond shape near the top of the weapon before looping over again.

I’m not a militarily-minded person, but isn’t it counter to a soldier’s goals to have anything blinking, glowing, or pinpointing the soldier’s exact midline to enemies, much less something that does all three at once, and in red, the color that travels the farthest in atmosphere?

What was that about “discreet” interventions?

Gestural disguise

29 Apr 2013 by Christopher Noessel

When the Mangalores meet with Zorg to deliver (what they think are) the stones, their leader Aknar is wearing a human disguise. The exact nature of the speculative technology is difficult to determine. (In fact, it’s entirely arguable that this is a biological ability, but it’s more useful to presume it’s not.)

Zorg tells Aknar, “What is that you? What an ugly face. It doesn’t suit you. Take it off.” Aknar strains his chin upward and shakes his head rapidly. As he does so, the disguise fades to reveal his true face.

Presuming it’s a technology, the gesture is a nice design choice for the interaction. It’s not a gesture that’s likely to be done accidentally, and has a nice physical metaphor—that of shaking off water. The physicality makes it easy to remember. Plus, being a head gesture, it can be deployed in the field even when carrying a weapon. This makes it possible to dismiss and show your identity to comrades without the risk of lowering your guard. It does temporarily limit the wearer’s ability to sense danger, but I suspect Mangalores care more about keeping their finger on the trigger.

Of course it raises the question of whether what results from the shake is just another disguise, but that would depend on some external system of multifactor authentication that’s separate from the gesture.

Roach Cam

25 Apr 2013 by Christopher Noessel

To learn the plans of the President, Zorg’s flunky named Right Arm infiltrates the briefing room via a remote-controlled cockroach. This adorable insect has a small parabolic receiver antenna on its back. Right Arm can watch what it sees with its eyes and listen to what it can hear through its… cerci?

The screen he uses is mostly full of the roachcam video. But it is unfortunately surrounded by some screen-green sciencey nonsense. A row across the top is headed “001” with rectangles labeled “MOVE”, “METHOD”, “CHECK”, and “SYSTEM.” A row just below is headed “A-B” with rectangles labeled “SPEED”, “TIMER”, “EXIT”, and “FILTERS.” A column of screen-green nonsense shapes fills the left side. A small butterfly-shaped graph at the bottom-left is labeled “CHECK.” A small box labeled “CALCULATING” is in the lower middle, which occasionally fills with scrolling nonsense. The right side of the screen is full of a circular graph showing a seizure-inducing flashing pair of green concentric circles. A green 8×8 grid on the right stays empty the whole time, though it is arguably the most likely useful thing, i.e. an overhead view of threats, say, like presidents brandishing cockroach-smashing shoes. Below the unused grid is a diagram of the roach itself, probably useful for understanding the health of the vehicle. Below that is a bit of unintelligible text reversed out of a gray background. When the roach nears the President, a bit of green nonsense text appears overlaid on the video feed, though it never changes.

I think this screen would have been less distracting and more helpful for Right Arm if you stripped away all the gunk at the top, the nonsensical overlay, trashed the column of hastily-drawn icons, saved him the constant distraction of the seizure circle by removing it, and leaving him with the two things that would actually be useful: the map (populated of course with some useful information), and the roach health diagram. Even though this screen is seen only for a few seconds at most, it reads as if it was hastily put together, unlike most other things seen in the film.

He is able to control the roach’s movement by means of a joystick with a rotating head. In contrast to the screen, this provides exactly what Right Arm needs to control the roach, and no more. He can move it forward and back by pushing the joystick forward and back. He can have it strafe right or left by moving the joystick appropriately. And when he wants to have it turn right or left, he can twist the joystick head in that direction. Pushing or twisting farther results in more motion. All told, a perfect input control for the task at hand. At least until you ask the roach.

Zorg’s desk

24 Apr 2013 by Christopher Noessel

When Zorg begins to choke on a cherry pit, in his panic he pounds a numeric keypad on his desk, clearly hoping that this will contact someone or help him in some way. His clumsy mashing instead causes a number of bizarre things to happen around his office; i.e. the doors lock, a lifejacket inflates (bearing the charming label HEAD THROUGH HOLE), a cactus raises and lowers, a Rolodex of photographs appears and spins wildly, a rack begins to shoot plastic wrapped tuxedo shirts into the air, cards spit out of a slot, and a strange piglet-sized, hairless pet with a trunk is roused from its napping place as it raises to the surface of the desk and stares at Zorg helplessly.

In the talk I give about the lines of influence between interfaces in sci-fi and the real world, I cite this as a negative example of affective computing.

If you’re unfamiliar with it, affective computing largely deals with giving computers a sense of emotion or empathy for their users. In this case, of course Zorg doesn’t want to summon his elephantito from its adorable genetically modified slumber. He’s panicking. He wants help. The joke in the scene is largely about how the unfeeling technology on which Zorg relies is of little practical value in a crisis, but we know that a smarter design would have accounted for this case of panicked mashing.

If (a bunch of key chords are pressed rapidly in succession) {summon help}.

Interaction designers should take care to learn from this fictional example that though some scenarios may be rare, they may be dire enough to demand design attention.

This poor Ouliman Akaptan is named Picasso, designed by Hélène Girard.

For a general analysis, I find the number pad to be the worst choice of input for this system. On the plus side it’s useful for arbitrarily-long combinatorial and chorded input. It’s for this reason the telephone network system adopted this strategy to provide access to any one of its 10,000,000,000 nodes. (And that’s only with a ten digit number.) Fine. If Zorg needs a phone pad for dialing numbers than give him a phone. But for this desk interface, it burdens his long-term memory, forcing him to remember the codes for the things he wants. If he really has only around a dozen or so things to control, give him individual controls that are well grouped, distinguished, labeled, and mapped. Also in taking this tack, someone in his service might have thought to give the vengeful, psychopathic industrialist an actual panic button.

Floor sweeping robots

22 Apr 2013 by Christopher Noessel

To illustrate his capitalist ideology, (a high-tech version of the parable of the broken window) Zorg activates his automatic cleaning robots. To do this, he deliberately crashes a glass to the floor, where a set of robots come scuttling out from beneath his desk and begin cleaning up.

Three of them serve to demarcate the space as a ”robots working” zone, with tall masts from which red beacon lights warn anyone nearby. In the middle of these three, a sweeper robot gathers the large pieces of glass with broom and dustbin actuators.

Then, a vacuum robot spins above the location to remove the fine pieces of glass.

Finally, a hemispheric robot also comes to sterilize the area, or possibly to just spray a pleasing scent in the air. After they are done, they retreat automatically to the desk, and a new drinking glass rises from a hidden compartment to Zorgs desktop, filling with water to the accompaniment of a small voice that announces “water” and, as a bowl of the stuff also raises, “fruit.”

As Zorg pulls a cherry from the bowl, the same voice announces somewhat pointlessly, “a cherry.” (Perhaps useful if the eater is unfamiliar with basic types of fruit.)

Analysis

The robots are meant to do their job safely and efficiently, and then put themselves away as quickly as possible. The main “interface” task they have is to keep nearby humans informed and safe. (Did Asimov write a law for vacuum cleaners?) A minor secondary goal might be to distinguish the function of each by their shape. The robots inform observers explicitly with the stanchion robots’ beacon lights and bright red patterns. In addition, the whirring sounds of each robot’s motors and actuators help to reinforce the fact that they are working. If they were completely silent they would be more problematic for people not looking or unable to see. The beacon might be a bit of overkill and distracting to someone at a distance, but since the robots are small enough to be a trip hazard, and Cornelius is in fact less than a meter away at the time of sweeping, I can see why it might be needed in this particular case. That they are each readily distinguishable means it might be easier to intervene or select a particular one for maintenance. So, aside from the faulty logic they’re meant to embody, mostly really well designed.

The main improvement I can imagine is that the system might reduce the trip hazard by unifying these disparate functions in a single device, and then either keeping them stanchion-high or flattening its top out like a step. But then we’d just have invented Roomba five years early.

Extra credit

This isn’t the first floor cleaning robot seen on the silver screen. There was another movie over three decades before that included one, even though it wasn’t what most people would consider sci-fi. Can any of the Make It So readers identify that film in the comments? (I’ll post pics of the answer once someone guesses it.)

The Chanel eye-makeup-erator

19 Apr 2013 by Christopher Noessel

After David offers Leeloo some clothes, he also offers her a device for applying eye makeup. Leeloo only has the most rudamentary grasp of English at this point, so to demonstrate its use he holds it up to his eyes.

This is a clear enough signal for Leeloo, who puts the device up to her eyes like a large pair of sunglasses. She can feel the momentary button near her left fingertip and presses it. In response a white ring around a Chanel™ logo illuminates for a second. Leeloo feels an unfamiliar sensation and pulls her face away, and we see that the device has applied complete eye makeup for her.

Analysis

The industrial design of the device is brilliant. It’s sized to be slightly larger than the eye area and it has the right shape for someone to know where to place it. The activation button sits exactly where the user needs it, and with enough of a button-like affordance that even without looking she can find it and press it. The device is just heavy enough to encourage supporting it with palms, which provides a firm base to resist too much movement on activation (and thereby risking eyeshadow right on the eye). The shiny black plastic reads like a cosmetic object, and professional enough that you can presume it’s safe to use near delicate eye parts. The white ring is a simple cue for those nearby that it’s in progress and not to interrupt the user.

A minor improvement would be to improve that simple light on/off to a progress ring that swept around. This would gave a sense of how much time it will take and how much time is left, even if it’s only a second.

The main question of the device is of course how does Leeloo specify the details of the makeup. A quick Google image search shows that the number of parameters is…um…vast.

Of course, if the device had some kind of low-level artificial intelligence, that agentive algorithm could handle a lot of the complexity for her, deciding on the best match for her schedule, fashion trends, current outfit, and her preferred position in the fashion-aggression spectrum. (Would there be a device that went up to 11 for drag queens?) But, when the agentive algorithm got it wrong, and Leeloo wanted to override those settings, she’s back to needing to tell the device how she’d like to override its suggestions. How does she do that?

Which raises the question of those three buttons across the bridge of the device.

Those three buttons

Of course three momentary buttons aren’t enough to control all the variables in eye makeup. Even if these are dials that control three variables, three variables aren’t enough. (Even if they were dials, why would they look identical to the momentary buttons? Things that behave differently should look different.)

Even if these buttons are not controls for variables but rather presets for sets of variables (Such as: “Work,” “Formal wear,” or “Defeating ultimate evil”), they’re not signaling their state well. Looking at that screen grab, can you tell which one is currently selected? I can’t. It should be apparent at a glance, so no one accidentally applies “clubbing” makeup when they mean “funeral.” So there should be some indication of what’s currently selected. Note that a lit button is not enough. Some descriptive text is needed. Such text would ideally be on both the “inside” and the “outside” so no matter how it was lying on a dresser, its state could be read.

Anyway, since those buttons aren’t sufficient for setting up the eye makeup, let’s hope that it’s networked to some other device with a richer interface, like a voice interface or Cornelius’ WIMP computer, where she can have a rich interaction for setting up those buttons.

With all that in mind, here’s another comp to illustrate these ideas. Admittedly, Chanel’s brand police wouldn’t be comfortable with an LED font, but it would clearly communicate that the text represents a variable and not a product name.

Special shouts out to You Yeti! who corrected our tweet that the two Fingernail-o-matics were not the only cosmetic interfaces in the survey. I love writing for a sharp, eagle-eyed audience.

Sci-fi interfaces

Stop watching sci-fi. Start using it.

Category Archives: The Fifth Element (1997)

Police light

It’s a terrible idea because of its placement

It’s a terrible idea because it’s a giant, glowing target

Improve it with a little bit of smarts

5E-opedia: Search

Search

5E-opedia: In-depth topic

In-Depth Topic

It’s still about the learning, stupid

5E-opedia: Watch and learn

Watch-and-learn

The Tyranny of Pause

Um…learning?

ZF-1

Analysis

Gestural disguise

Roach Cam

Zorg’s desk

Floor sweeping robots

Analysis

Extra credit

The Chanel eye-makeup-erator

Analysis

Those three buttons