The New Normal

Late June 2043.

The City of London.

Tap tap tap, tap tap tap tap tap. Mike furiously punished his backspace key, getting rid of everything he’d written aside from “Dear Elizabeth.” It wasn’t the first or even fifth attempt he’d made to write this email, an apology for abandoning their plans after the rally. As the crowd had turned on her he’d felt extremely awkward, and suddenly began to worry he might even run into some of his co-workers, or some of Elizabeth’s friends, or even Cassandra. So, he’d walked away. It was for the best, he’d written in at least three of the draft apologies, because he assumed she wouldn’t want to worry about having to keep their plans after going through all that on stage.

The reality was also that he couldn’t stand the sanctimonious bollocks. The moaning about automation, the moaning about the government, the over-wrought chants and endless marching back and forth. Coming from her it was particularly galling. He could still remember watching her defend it on TV. And then there was the matter of her family money. She raged against the very machine she built, while existing in a totally parallel world where the machine had no power over her whatsoever.

Meanwhile, Mike was finishing up his last week at HumanTouch.

He’d been lucky enough to get kept on to assist with the transition, so most of his colleagues had gone already, having been cut weeks ago. This sort of thing used to be a blessing. When he was younger he’d worked at a company which conducted a mass layoff, and he’d gotten a nice cushy severance package along with 90 days of what they referred to back then as “garden leave” – where the company essentially paid you to sit in your garden at home, finishing out the rest of your contract on full pay.

After a flurry of reforms, the nature of which Mike couldn’t really explain, that sort of thing simply didn’t happen anymore. No severance pay, no garden leave, no unfair dismissal complaints, nothing. As long as the company used one of the government-certified dismissal algorithms to determine who to let go, to ensure complete fairness, they were free to offer you nothing and terminate you as soon as they saw fit. The firm acquiring HumanTouch, Magenta, happened to have one of those close to hand. Because they happened to have built and released the first tool to ever receive that particular government certification.

Still, Magenta weren’t foolish. They wanted to retain as much of the firm’s existing business as they could. So they picked a few of the longer-tenured, higher-performing designers to stick around and ensure their AI models were appropriately trained.

Mike actually found the process fascinating. At first they’d simply let operations continue as normal, taking in requests from their clients, passing them to the designers, iterating on the results and then releasing them as normal. But they were, alongside all this, also feeding the inputs and outputs from each step into artificial intelligence models. And over the course of a few weeks, they’d see how close the AI was getting to the human output using the same inputs from clients. It was never perfect, of course, and the human designers would occasionally turn in something Mike might call “inspired,” but his new manager would certainly call “an aberration.”

By the beginning of the third week, they were quietly handing the output from the AI out to the clients, then having the human design team work through the feedback. By week five, the design team was simply supervising the outputs of each AI step, flagging anything that seemed odd, or waiting in case a client threw up a massive complaint.

None ever did.

By week six, the few remaining designers on Mike’s team, along with Mike himself, were “on standby.” They no longer had any visibility into what was going on, but were told they’d get a call if they were needed.

So here he sat at 3 p.m. on Thursday, the last work day of the sixth week since they were acquired. No calls had come in all week. It had given Mike plenty of time and space to draft, re-draft and re-re-draft his apology to Elizabeth. But it also probably meant Mike knew what the one call he’d end up getting would be.

Funnily enough, it turned out he was wrong. 5 p.m. came with nary a whisper. No phone call, no email, no nothing. “Huh,” Mike muttered to himself as he gathered his things and tucked them into his backpack. Each week had been a smaller and smaller group gathering at the pub for post-work drinks, and he feared today would be down to a single digit number of people. But he’d give it a go anyway.

On his way out of the building a rather unfriendly looking man in a black suit approached him just ahead of the gates, a small tablet in his hand. “Michael Fernsby?”

Ah, right. “Yes, that’s me.”

“Mr. Fernsby, Magenta thanks you very much for your service. Your access has been revoked, I can let you out.” Mike’s phone vibrated in his pocket. That’d be the notification of his building pass being remotely deleted. Right.

He walked into their usual pub at around the usual time, though with no colleagues in tow, and no colleagues in general. He briefly questioned whether it was the right place, given it was practically empty. Their usual bartender spotted him and waved. “Thank christ, something to do.”

He sipped his pint in silence, head on a swivel, wondering when or if anyone else he knew might walk in. It had apparently been dead all afternoon, he’d been told. Mike warned him the rest of his HumanTouch colleagues had been let go today, including himself, and the bartender just sighed and nodded, then poured a couple of shots.

Now though he sat perched at a table alone, half a Guinness in front of him, his phone occasionally lighting up with another angry or dejected message from their “Work mates” text group which he noticed had now been cleverly renamed “FUCK MAGENTA.” He picked it up and muted the notifications without reading any of it. About half-way through his first pint, he was still angry. But now he sat here halfway through his second a determined man. This wasn’t something he was going to moan or cry his way out of, he’d resolved to himself. He knew he was up against it, being a bit older, being from a profession that had mostly been automated away. But it happens. It’s happened time and time again throughout history. You can either let it beat you down and give up, or you can find your place in the new normal.

Mike had decided: he’d find his place.

But first, maybe one more pint. He stared at the news, on as usual for the City crowd, if only they’d showed up. His thought drifted back to the apology to Elizabeth. Maybe a text would work better? She probably had the same number. It didn’t carry the same gravitas as a letter, of course, but then he wouldn’t have to think so much about the contents. He was increasingly stuck on this, though. He had to get past it, get it out of his mind, so he could focus.

And then there she was on the TV. Mike’s eyes went wide. A photo of her from some recent rally, it seemed, with her fist in the air, in that same damn blue blazer she’d been wearing for half her life. “FORMER MP WILL STAND FOR, LEAD NEW PARTY IN UPCOMING ELECTION.” Mike shook his head and downed the rest of his Guinness accidentally.

God, why would she throw herself back into this? It was honestly so nice to see her after all these years. They still had a decent rapport, some good banter. And she obviously didn’t need to work. She had plenty of money. It was something he’d never understood about her, even as he’d try to encourage her whenever possible.

At least knowing this made the text easier to write. He grabbed his phone, dismissed all the nonsense, and popped open a message to Elizabeth.

“Saw you on the news,” he wrote. “Picking up right where you left off I see. You look great as always. Sorry to have missed our plans after the rally, and most of the rally itself. Wasn’t feeling well. You’re probably right about the panic attacks. Hopefully drinks and proper catch-up soon?”

He tweaked the punctuation, swapping one period for an exclamation point, then another. Then both back to periods. Then a mix. Should this end with a question? Or should he declare ‘drinks soon!’ He decided to stick with the question.

And, send. That struck the right balance, Mike felt. Casual, apologetic. Admitting to her that she’s right. He felt good about all of it. He turned to pop back over to the bar for another pint when his phone vibrated. Wow, instant response? He looked down and frowned:

“Your message did not send. Nobody with this username exists or their privacy settings prevent you from contacting them at this time.”

Okay. Great. So it turned out he didn’t have an easy way to reach her after all. Had she blocked him? It’d been nearly 15 years. Maybe she just had a new username. Or a totally new app, to be fair. So much had changed since then. He could try to find her on Popround, or … no, he couldn’t spiral. He wouldn’t let himself spiral. He got another pint.

Right. Starting his new life in the new normal. First thing’s first: a job. He’d found the job at HumanTouch through an old university friend who already worked there. It had been a while since he’d tried to find anything. And he was up against hundreds of thousands of others. But he was confident in himself.

He popped his earbuds in and opened YouTube. There was no room for shame anymore. Time to get some help. He searched “how to use ai job apps.” Much like dating, applying for jobs had changed dramatically in the past decade. Luckily, in addition to being a vast archive of older videos, YouTube was still haunted by people his age, still happily churning out how-to videos, a community of people helping each other survive and thrive in such a rapidly changing world. It was worth the time and effort to get the offshore network licence.

He found a pleasant-enough video called “Best AI Job Application App I Could Find (Britain),” made by a Cornish woman with the channel name “LyndseyHelps” who looked to be in her mid-50s. She had curly silver hair, green eyes, a warm, soft voice, and was sat quite close to her camera in what appeared to be a darkened sewing room.

She spoke at length about what had led her to need an app in the first place. Her husband of 30 years had passed away about five years ago, and he had historically been the breadwinner of their household. They’d always lived rather hand-to-mouth. She had no savings. And she refused to give up the home they’d shared for so long and be relocated through a government programme, which meant she didn’t qualify for any benefits. So she’d thrown herself head first into what she knew was probably a futile job search.

It was a moving story, but it reminded Mike of why he hated having to watch how-to videos.

Anyway, nine minutes later, Lyndsey explained that she really liked an app called Jobsworth. Pretty soon Mike was set up on it as well. To her credit, it was extremely straightforward: the app would ask him a few freeform questions about himself, what he was good at, his experience, then it would generate and send hundreds of job applications instantly. It would continually do this any time an opening matched his experience and interests, then it would let him know once it had found something. It seemed almost too easy, and he wondered why everyone wasn’t also doing this. It promised to let him know within a week what it had lined up for him.

Of course, none of this was free. He’d paid £8.99 for the Jobsworth app and would pay 5p per application it made on his behalf (with the option to set a budget per day, per week or per month). There was also a £4.99 per week subscription regardless. He figured this was time sensitive, and 5p was nothing, so he didn’t set any budgets – Lyndsey had strongly suggested that not setting a budget led to the best result.

So, the new job was sorted, it seemed. He just had to be patient. He flicked through the ever-growing list of applications it was sending on his behalf, all of them going to “Pending,” then a handful going to “Screened,” and a few of those going to “Rejected.” It was all extremely fast. Bots talking to bots, he assumed. Criteria being assessed. As best he could tell, the hiring and firing at many large firms was all fully automated. While he thought about how a system like that might work, a few more of the “Pending” applications flipped straight to “Rejected.” Hmm, it did say to wait a week or so. Better not watch how the sausage was made. It might get demoralising.

He closed Jobsworth and mindlessly flipped over to Linkup. He’d had a smattering of new matches, and it was once again offering him some auto-generated tips on how to improve his chances. He’d disabled every bit of automation when he first signed up, and was proudly a non-paying member, all of which it was happy to make clear to others if they looked at his profile. Shameless, absolutely shameless.

This wouldn’t fly anymore. Mike the graphic artist at HumanTouch, drawing everything by hand, that’s the guy who handled his dating profile like it was 2015. If he was going to live in the new normal, he had to fix this.

Back to YouTube he went. This one took more effort. His fingers hovered anxiously over the keyboard, the cursor blinking ominously in the search box. He could barely bring himself to type it.

“how to use linkup ai dating”

And off it went. He flicked through a few pages of results almost without looking, unable to bring himself to see them for fear of dying from embarrassment. He snuck a quick look over his shoulders to see if anyone could see what he was doing. All clear. Still empty. The bartender caught his eye and nodded toward the taps. Mike nodded.

Fine, fine. He had to just get it over with. What was there to lose? His eyes went to a result toward the middle of the page, “Best AI Dating App I Could Find (Britain).” Huh. He tapped on it, and was once again greeted by the warm, soothing presence of LyndseyHelps, in her dark sewing room somewhere in Cornwall. She started talking once again about losing her husband, in her hushed, even tone. Needing to put herself back out there, not knowing where to begin. It was genuinely heart-breaking.

Mike found himself thinking about her, and not paying much attention to her tips on how to configure Linkup. He couldn’t help but picture her sat there all alone in her marital home, widowed, working to support herself, trying to find someone new to spend her life with. And still finding time to record these how-to videos. Sharing her experiences with the world.

The bartender set a fresh pint down in front of him, breaking the trance. Mike took a quick sip and paused the video, then switched back over to Linkup. It popped up another offer of help with his profile, and this time he simply tapped “Yes, please!” A virtual agent appeared on screen and began asking him questions about himself. Seemed easy enough.

Eventually he was facing down a screen with a list of tiny, round pictures of faces, names, ages, locations and “suitability scores.” Next to each one had a status: Pending. After a few minutes, a handful changed to “Screened.” He scrolled down a bit and saw a smiling face that piqued his curiosity. Penny, 44, Wandsworth. She was quite cute! And she had an 84% suitability score. That seemed promising!

The status next to Penny changed from “Pending” to “Rejected.” His thoughts wandered back to LyndseyHelps. He wondered what she was up to right now.