Life as we know it is about to be transformed. Jobs, as we know them are about to be automated.
These changes are not far away into the future. They are imminent. About to take place in the next two decades. From 2020 to 2030, and between 2030 and 2040.
The purpose of this article is to scare the reader out of complacency and settling in the comfort zone of blissful ignorance. There is literally nowhere to run and hide from the coming waves of job-killing automation. The narrative is meant to drive you, the reader, to the point of realizing the need for a Universal Basic Income (UBI) when jobs (the primary means of distributing purchasing power) are not widely available for humans.
The list below is not exhaustive, it is merely an example of how automation is affecting human jobs from the low skill jobs to the high-skill jobs that require college education.
It is no secret. Cleaning is a demeaning and boring job that can be done by robots. A simple google search yields 1.5 million results. The technology is getting cheap too and the cleaning robots are getting sophisticated to handle more than just mopping the floors.
2. Farm Laborer
Grain farms have been automated for a long time in terms of tractors and combine harvesters. Things are about to take another turn for the worse for farm laborers. Precision farming, the use of drones, and harvesting robots are getting to a level where the entire farming process does not need any human labor. In the vid below, the robot uses a camera equipped with excellent lighting to identify ripe pepper riding on computer vision capabilities and cuts the pepper, and drops it into the basket.
3. Cashier and Bank Teller
Banks are gradually eliminating banking halls and ATMs as we move towards a cashless society. Nearly all retail banking services can be delivered via apps and remotely.
TymeBank, in South Africa, opened 2 million accounts within a year of launching. The bank has no brick and mortar branches and has no tellers.
Amazon’s much-hyped Amazon Go supermarket has no checkout lines.
It has no cashiers. Amazon automatically charges your account when you walk out of the shop.
Cars can drive themselves. You probably noticed the resistance, wailing, and noise from metered taxi drivers when Uber stepped into the transportation markets in major cities.
Now imagine the furore when Uber replaces the human-driver with a self-driving car. The long-distance truckers too can be replaced with self-driving trucks. This is not going to happen in the future. It is happening now. Wake up and smell the coffee. The technology is already here, once legislation paves the way like it did in Texas, it's a done deal. The video below is one of the many examples.
5. Cook/Chef/Kitchen Assistant
Cooking is an art, right? Is it? Or is it rather a science? Can it be automated? Can we also automate food preparation and the actual serving of food? It turns out, tech can solve many food-related problems from farming to harvesting, from food transportation to storage, from food preparation to cooking and dishing out. The only part that needs human input is eating.
The globally popular Instant Pot is an example of a small level of automation.
Meals can be 3D printed, although this technology is in the infancy stages, it will eventually mature and be embraced by society. KFC is working on 3D printed chicken nuggets. See vid below.
Currently, food supply chains up to the point of consumption are being automated. The entire restaurant can be automated, as shown in the video below.
6. Petrol Attendant
Most gas stations in the US had been automated with customers serving themselves. In Africa and Asia, working at a gas station (filling station, garage, service station) is one of the most basic entry-level manual jobs.
Electric vehicles will decimate the jobs of petrol attendants. EV charging stations will be fundamentally different from fossil fuel filling stations.
It's the end of an era for fuel service stations. It's an end-game. It's checkmate. Game Over.
7. Automotive Mechanic
As the self-driving electric vehicles roll in, automotive mechanics will gradually be replaced by robots. Most EVs will be manufactured by machines in fully automated factories. The servicing, repairs, maintenance will be performed by automotive robots too. You cannot take these vehicles to your mechanic down the road, no matter how skilled he is. It will be too expensive as opposed to getting back to the factory for repairs by robots.
Diagnostics are already performed by a computer via the computer box and the 60 to 200 sensors that are everywhere in vehicles in use already.
EVs will definitely put the neighborhood mechanic out of business. The usual narrative that repair mechanics will evolve and adopt is misplaced due to naive extrapolation of the past without digressing through the factors at play.
As usual, most auto mechanics have their judgment clouded by denial and the massive cognitive dissonance associated with pondering on the future such that they do not realize that automated centralized facilities will soon take over and that their role in the industry is dying.
8. Factory Worker (any factory)
Factories are getting fully-automated. Factory automation is nothing new. From the Industrial revolution days to the assembly line days to this day. Its technologies’ relentless march. There is no going back.
The video below shows an example of a fully automated factory. Humans are only needed to monitor and maintain the system.
9. Tele-marketers and Customer Service Officers
When automation is mentioned, most people limit it to jobs that need manual labor. As it turns out progress on the Machine Learning (ML), Artificial Intelligence (AI), and Natural Language Processing (NLP) fronts are set to automate office jobs too.
The Call Centre sub-industry is getting automated very fast. NLP successes, aided by neural networks and Moore’s Law is advancing the capabilities of chatbots. Very soon humans will prefer messaging a chatbot than an actual human being. Why? Because a chatbot will be able to answer them better and faster leveraging unlimited answers in the Artificial Neural Network (ANN) developed from millions of real-life examples. An inflection point is imminent.
A call, answered by an automated bot might be more satisfying to a human, and the human might not notice that he was talking to a bot. The bot can recognize the voice, memorize facts, customize the response taking into consideration the millions of data points that it knows about the client.
The video below peeps into the process of developing better chatbots.
10. Paralegals/Legal Assistant
We are going up the ladder of occupations.
Library searches, document review, case planning, legal research, summarizing of documents, and fact-gathering are some of the functions that can be done well by algorithms. These tasks can be carried out by software and algorithms.
Lawyer Ross, legal research software is essentially an AI robot that undertakes paralegal work. It's built on IBM’s Watson program that beat human competitors in the TV show Jeopardy.
Ross will cost the firm $69 per month. This goes a long way in replacing paralegals.
The AI in use is like a google-search on steroids and customized for lawyers.
Some say lawyers themselves will be replaced by bots too as bots learn and master human natural language. But that is not imminent.
11. Data Capturer/Office Assistant/Bookkeeper
By now, almost every other mid-sized-to-large office in the world has seen some benefits of automation. The switchboard was automated a long time ago. Paper trail is being eliminated with most offices endeavoring to be paperless.
Data capturing is being automated as well. Office Assistants, Data Capturers, and Bookkeeper jobs are by and large data capturing jobs. These roles involve capturing data 60–70% of the time.
Digitization, and software integration, are eliminating the need to manually capture data. Forms are being sent, populated, and signed online with the data flowing into the system (accounting system, practice management system, booking system, or whichever system in use) thus eliminating the need to capture data from printed material.
Softwares can communicate with each other and send and receive data thus eliminating the need to capture data from one system to another. The API is eating the Data Capturer for breakfast.
When you mention the Bookkeeper for jobs up for automation, the Accountant thinks he is safe. Well, he is not.
In this area, just as in many other white-collar jobs, human workers think they cannot be replaced by machines because of technological ignorance and sheer denial. Most accountants are familiar with “machine learning” as a buzzword but do not understand what the process entails.
Accounting is a perfect example of a job that can be eaten by software. It is rule-based. The rules are static and rigid. The input data is usually clean and is of a consistent format. The example data sets are huge, they exist in millions and they are all digital. The input data is already labeled. The desired output (target) is known, accountants have been producing the desired output for decades.
This field is ripe for Machine Learning, supervised learning to be specific. General Ledger Account allocations are fodder for a Classification Algorithm.
Xero, a cloud-based accounting software, is making use of machine learning to “learn” how your business is run and automates repetitive tasks. It can intelligently “guess” account allocations, populate invoices, create bills for you, create and send invoices to customers (repeat templates), automate bank reconciliations. It's only a matter of time before the software becomes entirely reliant on machine learning capabilities and does not need you, the accountant.
Xero is making use of ML to choose the right GL account codes using your previous transactions. However, it does not really need to be limited to your company alone. he algos can learn from other companies too. There are more than 10 million different account codes in Xero. The ML algos should be able to sift through these and recommend the perfect account code and name, create it, and post transactions. This ability to learn from others (in their millions) means transactions don't rot in the suspense account.
Think of this, if a car can be automated in a way that it learns to obey the rules of the road, then surely accounting can be automated in a way that the system learns to obey the principles of accounts, the double-entry, the IFRS, IAAS, GAAP, jurisdiction-based tax accounting, trust accounting et cetera.
The plot is getting thicker and interesting. The radiologist will tell you that Artificial Intelligence will make them better, it's an excellent tool that is helping them to do their job. What many do not realize is due to Moore’s Law, things are about to change. Whereas technology used to be a tool in man’s hands that helped him work better, the tool has improved a lot to an extent where the tool can do all the work without needing any human intervention.
What do radiologists do? They study images and use their knowledge to interpret these images. The interpretation is essentially diagnosing various medical conditions. They make use of medical imaging tools such as X-rays, MRI scans, ultra-sounds, electromagnetic radiation. Furthermore, they use this to provide therapy, which is exactly where they are needed.
The diagnosis and the interpretation can be done by machines and the treatment can be done by humans, for the foreseeable future until the treatment can be executed by robots.
One neuroradiologist thinks that in 10 to 20 years, most imaging will be read only by a machine, and results will be interpreted and transmitted directly to the referring physician without input from a human radiologist. 20 years is a short space of time, a student entering college now to study radiology might only be able to practice radiology for 16 years after which he/she won be needed anymore.
“Stop learning radiology right now”, it will be replaced by Deep Learning algos.
The video below shows an AI-based system being developed at Stanford University that performs slightly better than human radiologists in diagnosing several pathologies.
What does a pharmacist do? They dispense prescription medicine and give advice on over-the-counter remedies for small health issues. Dispensing medicine responsibly is all about possession of knowledge of medicines.
It turns out that the body of knowledge stored in a pharmacist can be stored in the cloud and accessed when needed. And, it turns out that the physical act of dispensing “in the absolute sense of handing over medicine to a client” can be done by anyone, including delivery persons.
The job of the human pharmacist is dead. No need to say nice words for comfort. It is what it is. End game. Checkmate. Game over.
On the 17th of November 2020, Amazon Pharmacy was launched. It's the beginning of the end for the neighborhood pharmacy. Jeff Bezos is the richest man on the planet for eating America’s retail. Started off with books, then clothes and everything else. It was fun when Amazon’s recommendation algos and delivery systems were eating Malls for breakfast, mom-and-pop businesses for lunch, and Walmart for dinner. Now, it is coming for the pharmacists. You cannot run and you cannot hide.
Go to the website, https://pharmacy.amazon.com/ and watch the video, or sign up, and ask yourself if the neighborhood pharmacist will be needed any longer.
15. Risk Analyst, Insurance Underwriter, and Actuarial Scientist
For the article to be shorter, I have bunched these three jobs into one. What does a risk analyst do? Whatever he does is getting automated very fast. He will not be needed very soon. Claims handling and underwriting can also be performed better by complex algorithms.
Actuarial science is often regarded as a very difficult and complex area of study. It requires the best of brains to understand the complex statistical, analytical, and mathematical problems that are inherent in life, notably in insurance, investments, and risk management. Artificial intelligence is especially good at this. It can perform these computations better and faster than humans.
Actuarial Science is an endangered profession. A human being needs 6–10 years of training to be a proper fully qualified actuary. An algorithm only needs a huge repository of data to iterate over and over (i.e training) for it to become a useful algorithm. The daily calculation of reserves for an insurance company can be executed by a well-trained algorithm. With data available, training the algorithm can take 6–10 days.
Moore’s Law (relating to the processing power of computers), Metcalfe’s Law (relating to how the value of a network is proportional to the number of connected users), and Kryder’s law (relating to storage space for data) are leading the technology’s relentless march. Data collection points within a network setup allow the AI systems to collect more data than can ever be analyzed by a human actuary, the Internet of Things, and people with sensors everywhere will allow the AI systems to predict with an insane degree of accuracy where the next risk event will take place.
For example, the security tracker on your car can provide GPS data, travel data, location data, etc. about your driving patterns, driving speed, etc. This data can be supplied to your insurer by the tracker company. An AI system can analyze this data and provide a personalized insurance quote for you based on your updated risk profile. Multiply this across a million clients, you get HUGE data sets, BIG Data, that knows where the cars are, who is driving, and what speed, and who is parked, where. All of these data points can be fed into a Master Algorithm that calculates reserves, thus reducing the need for several actuaries.
Good journalism is art. This too is getting automated. Report writing software is already in use for writing sports articles from live matches and post matches.
Will Journo’s be replaced. Microsoft has an answer. https://www.bbc.com/news/world-us-canada-52860247
Drones can capture live footage. Reporting writing software can write articles just like how humans do. Well, the algos, are not yet at the level of fusing jokes and satire into articles but they can already pen a decent, objective account of an event without being overly influenced by the need to offer a narrative.
In China, they are already making use of AI Robots as news anchors. See the video below.
It is the end game for most human jobs. We can go on and on, looking at how each and every existing job is being automated in a manner that makes it humanless.
The Law of Unintended Consequences is interesting. If all human jobs are replaced by robots and machines, where will humans get the money to buy all the cheap goods and services offered by tech? The unintended consequence is the destruction of aggregate demand for the vast majority of humans.
A different mechanism, other than jobs, will be needed to ensure that the vast majority of people have the means to purchase goods and services.