Automated Incompetence

One of the threats that many people fear from AI and robots, is a threat to the workforce with AI and automation taking everyone's job. This is known as "technological unemployment", when someone loses their livelihood and income, due to being replaced by a machine that can do their job quicker, better or cheaper. 

I think I've made it clear in my previous posts that I don't think it's possible for a robot to be superior to a human being. Humans can outperform robots most of the time, although that does mean that tasks which require little effort from a human can easily be replaced or reduced by productivity-improving technologies.

I'm not saying that this means those jobs aren't significant or that those workers who lose their job to a robot don't matter, I'm simply saying that the abilities of our machines currently coincides with the ability of jobs requiring mostly physical labour.

In my opinon, it's better to risk a machine to dangerous or repetitive labour, however I also accept that these decisions are driven by an employer's capitalism not their compassion.

And robots can only do certain jobs well, meaning we have nothing to fear from robots when it comes to performing mental labour. Even our most advanced artificial intelligence (at time of writing) can't reliably perform these tasks anywhere near as well as a human.

So, that's it right? Robots can't take our jobs.

Unfortunately, no... Just because a robot can't do your job doesn't mean it won't.

Robots, for example, tend to write in stilted ways that people don't find convincing, yet they're already taking (or threatening) writers' jobs, being used for copywriting, advertising and content generation. Robots can't empathize - can't even feel emotion - but there have been robots made to assist nurses and caregivers and chatbots designed to use talk therapy for helping those with mental illness. Robots have no opinions, no emotions to express and no artistic style, yet they're being used to create digital art.

Yes, a robot can talk, listen or paint a picture, but that doesn't mean they do these things well - just because something has feathers doesn't mean it can fly. But when I said in the post titled Robots are Not your Enemy that robots aren't the bad guys, I meant it. Robots aren't the ones who want to take your jobs away from you - it's always an employer that wants to do that. And not all employers realize how flawed robots are...

Actually, as much as I want to believe in our collective humanity, and say that employers are wide-eyed idealists that don't realize the deficits of artificial intelligence... I don't really believe that for a second.

I think employers know that robots aren't great at these jobs, but they don't care. Employers who even consider the use of robots clearly don't value their employees work - if they did, they wouldn't see them as so easily interchangeable - so, why would they care that a robot does a poor job?

I believe that there are a lot of employers - hell, a lot of entire industries - that will gladly fire some or all of their workforce if it means they can save pennies on the dollar.

This is the future we're looking at, in the coming years - and it's not only going to happen, it is happening.

I'm a writer so I care about writing and the future of literature and art. Even though I had no love for BuzzFeed I was horrified to learn that they fired several writing staff, to replace them with AI. And I didn't want to believe that a writer would use AI to write their stories, but I was wrong. Clarkesworld, a venerable science-fiction magazine that would allow writers to submit stories to for publication had to close their story submissions in February this year, after their inbox was flooded with AI-generated stories. According to the editor-in-chief, Neil Clarke, over 1200 stories were submitted to them at once, and 500 were clearly written by AI. Their submissions are now reopened, but they've changed their policies saying they won't accept AI-generated stories, and anyone even suspected of trying to submit a story written with A.I. will be banned and blacklisted.

I didn't think writers would stoop so low, but it may not have even been writers. After all, Clarkesworld is a pro-rate magazine, they offer 12 cents per word, and wordcounts of 1,000 - 22,000 words.

That's a potential profit of $120 - $2,640, I can see why an immoral opportunist might try to take advantage of that, if they could - and AI provided an opportunity for these grifters to try.

Although again, maybe I am letting my bias persuade me... maybe there are writers who have no love for the craft.

So, is this the true A.I. Apocalypse? Rather than networks setting off nukes, kill-bots gunning down civillians and assassin drones hunting down resistance... society will instead slowly crumble as we replace competent workers with stupid-systems and idiot-bots that turn our manufacturing, media, medicine and military into a collection of idiots who occasionally confuse a pedestrian for an "unknown object", identify a ruler as a tumour or even mistake clouds for ballistic weapons.

Admittedly, that's an extreme example scenario, and I don't think the world will be destroyed by moron machines. However, I do think that this is our near future, and one that will get worse as time moves on if we're not careful. Already, we're seeing the growth of "A.I. Accidents", with a growing database of incidents and whilst I don't think this is the end of the world, if we keep giving jobs to A.I. that it cannot competently complete, then it is going to hurt us in one way or another.

In conclusion, I'm the Absurd Word Nerd, and I hope that these are merely teething troubles that we can resolve before letting A.I. take over too much of our lives. But, if we can't iron out these kinks, then I think it may be a sign that this A.I. revolution isn't truly a world-changing development. Rather, it may just be yet another technological bubble - a load of hype that will eventually burst in our faces.