Who Will AI Replace, and How Fast?
By Josh Bachynski, MA, PhD
The world is changing, they say. This truism we hear on a regular basis, however, has never been more true. We are literally on the cusp of a new era in human civilization. The era in which robots start doing much of our work for us. This will radically change society from the models setup in the last industrial revolution: the labour model. And as society reconstructs itself, those who work in areas of society about to be phased out, might do well to take note.
But the question is: who will get replaced first? And when? Who should be looking to change careers in the next 5-10 years?
In this short article (which is a snippet of the book I am writing on the subject) I will try to give notes as to what to expect in the next two decades.
But in short, a sure-fire way to know whether your job will be replaced is ask yourself this: Can a trained monkey do my job, or most of it?
If that thought has ever crossed your mind, then yes, you are in great danger of being replaced in as little as 3 years.
Here are some other quick things to think about when thinking about how the oncoming robot job apocalypse will affect this world:
The first rule of thumb: The First-world will be replaced first, the second and third have cheap labor, and low buying power. So they will not implement robotics or AI much, until it becomes dirt, dirt cheap. Say 20-30 years.
By this I am not recommending first-world laborers move to the second or third world to find labor jobs, for the living conditions in these areas will only be made worse by climate change and economic depression from first-world countries who previously outsourced moving manufacturing back to the first-world with robotic workers. Also remember many geographic areas that think they are part of the first world will become the second and third due to climate change ruining their living conditions. I’m talking about areas in Europe, US, Australia, and Canada, etc. Any article which predicts the future cannot do so in a vacuum, and must take into account the current climate. And as we well know by now, the physical climate is a changin’ and along with it the economic and social and psychological climates will also become harsher.
Also, another rule of thumb: the more the currency value of the country is high, and the higher cost of living increases, the faster they will and must develop cheaper labour solutions, and unless outright slavery is once again made legal, they will need AI / robotic solutions to replace or supplement their labour force (not to mention to outcompete China. The only way the first-world defeats China from becoming the de facto economy and thus power on the planet, is to take their revenue sources away from them – and this means taking manufacturing away from them – with the first-world’s high cost of living, this can only be done by more efficient AI and robotics).
Another rule of thumb: when will it be cost effective? For the AI solution to be implemented, the total net cost of the AI solution (including health care, recruitment, liabilities, etc.) must be less over time than having humans do the work. AI and robotics will not be implemented until this is the case. So you might be wondering: how is that possible? Humans are so smart and adaptable. You can just talk to us, and we are trained. And we come with very dexterous hands. How could robotics ever beat us at that in the near future?
Well, consider this: within 5 years AI will be able to walk, talk, answer semi-difficult questions, and perform complicated tasks, with a reasonable battery life, with the ability to recharge nearby. AI can and will work non-stop, no breaks, no holidays, for 24 hours a day, faster and with less mistakes, so they improve productivity by 10% to 100% plus depending on the industry.
But also a major shift in society will occur. The very conventions set up to house and allow humans to work will no longer be required. So much cheaper buildings will and can be built to optimize work for robots. Indeed, the building itself will be a robot (3-D printed by other robots), with many little robotic drones inside to do the work faster and better than any hundred humans could do.
With drone delivery, “outlets” of any kind will be redundant. Indeed warehouses could be redundant. Why do we need a house to store wares anymore? When drone delivery chains managed by AI software systems keep track of every item and where it is on the planet. The wares are stored in situ, wherever they are, in robotic drones, travelling around the planet. Once printed from the factory, it can be sent directly to the consumer. All the contrivances of warehouses were only made to accommodate human fallibility anyways.
But that is more like 10-15 years off. At some point in the next 5 years the bugs will be worked out and the numbers will make sense to work in human environments. Ironically, in many industries, the robot will work alongside humans until it learns enough from them that it can do the work itself. So, as per usual, human labourers will train their replacements, and then be handed a pink slip.
The final rule of thumb: The rollout and implementation of these technologies will also take 5-10 years, so I expect the most vulnerable industries below to start being replaced in high-cost of living countries in under 5 years (hastened also by the “Great Resignation”), and take 5-10 years to phase out, with an utter maximum rollout of 20 years (and this would be in the event of some major first-world emergency occurs, like a severe economic depression, or the pandemic takes an even more lethal turn requiring severe lockdowns, or extreme weather events destroys the western or eastern coasts (in storms and or fires) causing a humanitarian and resultant economic disaster, etc.).
With that being said, here is the short-list of who will be replaced, by how much, and when:
Almost fully replaced: by software in 3-5 years or so
Receptionists, Hosts, Clerical Work, Phone Reception, Secretaries, low level Internet security and maintenance, education (in-home learning), etc.
Almost fully replaced: by robots in 5 years or so
Dangerous or Dirty Repetitive Work; miners, construction cleanup, garbage collection, city maintenance / cleanup, factory work, warehouse work, fast or casual dining food restaurants, food trucks, deliveries, taxi services, house cleaning, yard work, supply chain, low level watchdog security (terrestrial), parcel delivery, education (classroom), retail (low end items), groceries, etc.
Partially replaced (supplemented) in 5 years:
Too many to mention, but here are some standouts: Nursing*, medical doctors*, childcare, eldercare, police work, military, medical, retail (clothing and high-end items), video/audio/music/art creation and editing, software programming, accounting, the maintenance and watching of all fields not mentioned above – everyone will have a specialized AI tool to do something they do in life and work, if not many. Most people now already do, whether they realize it or not.
Let’s not forget wellbeing care or spa services (massage, sex work, chiropractic, acupuncture), etc.
*A note about medicine: there will always be nursing and doctor jobs and they will be supplemented by robotic nurses and AI diagnostics. Why? Because life conditions on our terraformed planet are only going to get worse. There will always be streams of never-ending sick humans to care for, and because humans are in the mix (the ones being cared for) robotic efficiency, and demand, can never surpass human help in this regard, especially if the nurse or doctor is willing to move. Doctors less so, they can advise and diagnose from abroad. Nurses here, the people who do the emotional grunt work, will always be in more demand. This will be phased out in 20-50 years as mentioned below when hospitals are sick people factories and or humans trust sufficiently cheap Westworld style androids enough to doctor and care for them.
And that takes me too…
What will not likely be replaced, soon if ever. Maybe in 20-50 years.
The following I have a hard time seeing being fully replaced, if at all. The main reason is this: humans are social creatures and need and want them (especially for things that look like us). We have no idea how much humans will like or trust these robots. Some cases: yes. Some cases : extreme violence and hate against robots. In the end, sometimes we will want to see humans and trust only humans, we will want to break the rules, negotiate, gossip, get opinions from someone who feels. AI is not well suited to this for a few decades. The more we need emotional connection, the less robots will work there until 40 to 50 years, and Westworld style androids that are indistinguishable from humans exist and are cheaper than the real thing.
I think industries like:
Real estate (too many questions, emotional connections), high end Car sales (same in so far as you trust your car salesman), retail clothing sales (need to try the clothes on, ask how it looks from someone who seems to care/be honest), jewelry sales (same deal), many counselling or advising what to do (life coach, wellness gurus; charlatanism notwithstanding), high end management, high end politics, high end art (with much subtext – the AI will have trouble with subtext for the next 20 years then AGI will become real, although very expensive and rare), etc. To make something pretty? AI can already do that.
Do not misunderstand, ALL of these industries, indeed every industry, will be bitten into and greatly supplemented by AI. AI will take over even some of the above work. But I believe industries with “soft-skills” will be the most resistant until humans are comfortable and trust AI more than people (given that our kids’ teachers will be AI this will happen in one human generation, 15-30 years), then these “soft-skill / human connection” jobs too shall fall. They have no more than 50 years utter max, maybe even as close as 25.
Conclusion: Over the next 30 years, in the West, we will all become managers in the AI system pushing humanity and or profits forward, luddite farmers and community builders (including ecological supplementers, running solar farms, putting back into the grid, etc.), and/or crypto workers building stock in crypto systems in the Metaverse (ready player one like) gaming system(s), and/or disaster clean up and humanitarian / medical work (due to unrelenting extreme weather), space explorers, miners, and colonizers, and/or human baby farmers (to replenish the human population, assuming we cannot solve the current fertility crisis). These will be the primary human roles due to our technology. Past, present, and future.
The rest of the world will suffer the majority effects of extreme weather from global warming.
To say the same thing another way, we all will suffer that. And never-ending pandemic. It is just the first-world will have the resources to maintain a certain standard of living through these increasingly harder times.
BUT first-world/second-world distinctions will also be eroded as well, geographically speaking, as long as sufficient broadband connection is possible people can work as crypto miners/stock builders in competing crypto systems in the ubiquitous Metaverse. Thus, language (the AI will translate all that on the fly, already can) and geography will not be effective barriers.
Welcome to the brave new world. What wonders await us in this undiscovered country, you will likely find out. Because, unless cataclysm occurs, it will happen faster than you expect, driven by the need and dream for innovation. If the only barrier to it not happening is “I cannot imagine how it will” well then it will happen: because someone can. There is more money behind it than God. It’s coming.
Prepare yourself.