“Now or never, Cal,” Howser says, hand on hips. The boy’s only got six months on Cal, but has grown lanky like a sapling and those extra few inches give him authority.
Cal stops chewing his lip. It’s important to show no fear in front of the others. Stay calm. You can do this. “Yeah,” he says.
He has to repeat it above the noise of the traffic. They’re standing on a grassy bank, wrong side of the security fence, looking down at six lanes of speeding vehicles just thirty feet away. Traffic’s picked up noticeably in the last ten minutes.
“This is nothing. Not if you’ve got guts,” Howser taunts.
I know. I’m going.”
Cal’s watched Howser do it. Now it’s his turn. Howser was first, of course. His idea, his chance to prove himself — and now it’s warped into some kind of initiation rite. The main thing is to walk out into the traffic with a purposeful, deliberate stride. That way the algorithms have time to figure out options: brake, change lane, speed up, whatever. Everyone knows driverless cars have reaction times a thousand times faster than humans. As long as you move steadily, their rooftop sensors see you coming and work out the danger. Algorithms and control circuitry do the rest. It’s a harder problem in heavy traffic with so many other variables — harder than a human could handle — but that’s kind of the point, he supposes. As long as he doesn’t hesitate, keeps moving in a predictable direction, he’ll reach the far side of the freeway, no problem. Howser’s already proved it.
What if there’s some jerk driving in manual mode?
Don’t think about that. Nobody drives in manual mode any more. Why would they?
“Hey, Cal! Not scared are you?” Howser asks, and the rest of the group laugh.
“Just waiting for the traffic to thicken a little,” Cal says. “I want to make this a good one.” The others fall silent; watching, waiting.
Cal lets out a long, slow breath and begins to walk steadily down the grass bank towards the speeding cars.
***
Luca’s sick of the endless road, sick of its monotonous rumble, sick of the metal/plastic prison of this car. And sick of this job. But it’s still a job — and when the factors are right, it pays well enough.
He snaps out of his reverie, noticing a string of compacts in the middle lane overtaking a lumbering tanker. They’re all models he recognizes, ones that run AutoRanger8.6 software — just the kind of target he’s been waiting for. Their corporate rivals pay handsomely for anything that dents AutoRanger8.6 safety stats. Anything. A shunt, a fender bender. It’s dirty and underhand, but in the safety business, reputation is king. It’s risky, sure, but bigger risks mean bigger payola. Unpredictability is the key to fooling the algorithms. Luca flips his car into manual.
That’s when he sees the boy in the road.
***
Cal knows he mustn’t look.
As long as he keeps going, he’ll be okay. He’s a visible, slow-moving obstacle, following a predictable path. Each vehicle’s control algorithms will recognize the danger, take avoiding action — and factor in their neighbors all doing the same thing. It’s a complex, real-time modelling problem but it’s all handled in the software. All Cal has to do is keep going.
Cal looks anyway.
He can already feel the bass rumble of the tanker barrelling north down the slow lane. Now he sees its bulk looming large. There’s a cluster of compacts spread across the next two lanes, with a red sports car weaving erratically in their midst, like a predator harrying the pack.
Every instinct is screaming, Don’t step into their path! Turn, run!
Somehow Cal keeps going, eyes fixed on the grass bank at the side of the freeway. One step, then another. Another.
A horn blares, deep and urgent.
A low moan escapes from Cal but he keeps walking.
***
The software in the tanker’s control module has already analyzed the problem, spotted the slow-moving obstacle ahead, identified twelve distinct scenarios. It’s prioritized the two most likely and conferred with a secondary processor to be certain of the analysis. Both control units agree on the necessary evasive actions. Handshake protocols with the surrounding vehicles are established wirelessly; data exchanged. From the tanker’s viewpoint, all but one of these vehicles are known quantities, their movements easily choreographed. The red sports car in manual mode is a tougher problem. It’s already behaving erratically. The surrounding vehicles reach consensus. If there has to be a compromise — a sacrifice — that unpredictable element must be top of the list.
The tanker swerves across two lanes to avoid the boy, moving into a space miraculously created as several compacts make violent, coordinated movements, much to the consternation of their passengers. But the red sports car has veered too. The tanker clips its rear fender, and momentum does the rest, sending it spinning into the central reservation. Its shiny bonnet crumples and the car flips over the barrier, wreckage spinning off in all directions — a new set of computational problems for the south-moving traffic which parts to make space for this intruder. The wreckage rams into a bridge support, barely recognisable as a sports car any more.
Luca regains consciousness only briefly. He can smell gasoline leaking somewhere but can’t feel his legs, can’t move. A great weight seems to be pressing down on his chest. He has time to think, I got the job done, didn’t I? Hell, yes! before the darkness comes to claim him.
The compacts speed on over the hill, unscathed.
***
Cal reaches the far side trembling, head full of the sound of tire-squeals and crumpling metal and the pounding of his own pulse in his ears. He keeps walking, up the bank over the low fence and into the fields beyond.
He’ll keep going. That’s the only way.
Steady.
Predictable.
Maybe he’ll even find a way to stop.
Eventually.
David Cleden lives in the UK, works in London and is the 2016 winner of the James White Award, with published work in Interzone, Betwixt, Electric Spec, The Colored Lens and other venues. His day job is writing business proposals but turns to writing fiction when darkness falls. He can’t stress enough how important it is not to get the two jobs muddled up. One day he will have a proper author’s website and write something intelligent on it, but until then he can be found on Twitter as @davidcleden.
Read EDF every day? Show us you care via Patreon.
