What Is Artificial Intelligence and Remote Working? What Is Its Impact Going to Be?
Historical Development
Unprecedented changes are happening in the world of work. What is of particular interest is, they are all happening at the same time.
We have the most significant advancement in artificial intelligence (AI) Figure 1.1, which is now working, and it is beyond our full comprehension. What is staggering—this has been happening dramatically since 2014. It is causing a massive paradigm shift that is irreversible, and it will change life and work, as we know it forever. To put this into context, if one looks at the development of humankind from 9000 BC to date, what we see in the past few years is expediential growth (Bauckhage 2017). The development of the gross world product has outstripped anything that has cumulatively happened before—and it has happened dramatically fast.
Three predictions that were quoted by Bauckhage:
By 2027 every process will be managed by A.I. (Bauckhage 2017, University of Bonn, Germany)
By 2027 70% of all S&P companies will have disappeared. (McKinsey 2016)
60% of all professions will be affected by automation. (McKinsey 2016)
Figure 1.1 Significant development steps
If you do not have a strategy for AI already, you need to activate one now. Progress is moving so fast that organizations cannot afford to wait; changes will not wait—for sure.
AI has come to the fore because of three significant factors.
1. The first is what we refer to as Big Data. We now can manipulate vast quantities of stored data and with these data can produce predictive outcomes. What’s been in the press often is, how our digital footprint is now being used. From the moment you ever switched on a computer, mobile phone, or tablet or used a credit card, data have been recorded, stored, shuffled, packaged, and sold by companies such as Axiom. Big Data is now available to all.
2. The next is affordable high-powered computing. Processing speeds and storage have increased, and the prices of computing have dramatically dropped. We now have quantum computers, some of which allow free access.
3. The final is the emergence of deep learning systems. These are systems that start to learn by themselves using cognitive learning—no need for old-style programming.
BD + HPC + DLS = AI.
You will hear the term neuro networks being used quite frequently these days; so, here is an attempt to explain what they are. In Figure 1.2, here we have the straightforward mathematical computation an input, weighting and addition, then that gives us a mathematical output.
AI, particularly deep learning, develops things further and has made rapid strides in a relatively short span of time.
In AI, we connect many layers of neurons; in fact, today we have millions of these as paired inputs and likewise a multitude of outputs. Deep neural networks are vast and very complicated; the big breakthrough that happened recently is that these networks now have cognitive ability to process; this has caused a dramatic improvement and change. It can be called self-thinking. The program automatically alters the weighting and keeps self-adjusting until it achieves predetermined outcomes Figure 1.3. The person credited for this is probably Geoffrey Hinton, the company who has been most instrumental in exploiting this GOOGLE.
To instruct AI and to get it to solve problems, we use algorithms. An algorithm is a detailed series of instructions for carrying out an operation or solving a problem. In a nontechnical context, we use algorithms in everyday tasks, such as a recipe to bake a cake or a do-it-yourself handbook.
Figure 1.2 A mathematical neuron
Figure 1.3 Deep neural network
AI system computers use algorithms to list the detailed instructions for carrying out an operation. For example, to compute an employee’s pay check, the computer uses an algorithm. To accomplish this task, appropriate data are entered into the system. What makes this particularly efficient is that, various algorithms can accomplish operations or problem solving easily and quickly.
However, it is important to note that a programming algorithm is not a computer code. It is written in simple English (or whatever the programmer speaks). It does not beat around the bush—it has a start, middle, and an end. In fact, you will probably label the first step start and the last step end. It includes only what you need to be able to carry out the task. It does not include anything unclear, or ambiguous (Study.com 2018).
There are breakthroughs in every area of business; the finance industry, for example, has found that AI is a better and safer way to do trading.
The Different Types of AI
When you decide to use AI, it is better if you confine yourself to your business needs rather than getting embroiled in the technologies that AI offers. Only then can you get the best results. For organizations, there are three main categories.
Process automation is probably the easiest to understand and is the automation of digital and physical tasks. These are typically back-office administrative and financial activities. Process automation could also involve robotic processes and automation technologies This approach is probably the least expensive and most accessible to implement as far a computing is concerned it is the most simple. This can be done using current technology now. In an example from Davenport and Ronanki (2018), NASA used process automation to look at its human resources (HR). In the HR application, it found that 86 percent of the transactions were completed without any human intervention. HR professionals need to pay attention to this advancement made by NASA.
Cognitive Insight
This is the second most common area of AI and the area that potentially will have the most impact on HR and the way that we work. This process uses algorithms to detect patterns in vast volumes of data and interpret their meaning. When this is coupled with deep learning, AI has the potential to take off to great heights. Cognitive insight is mainly referred to as deep learning, and in applications where this has been used, the results have been nothing short of amazing.
AI and deep learning are rapidly growing and expanding into every area of business. A qualified doctor, who is a specialist in cancer diagnosis, may take 2 or 3 hours looking at X-rays to diagnose correctly the symptoms that the patient may be suffering. Using deep learning, the same work can be carried out using AI in .03 of a second. AI is becoming more accurate than a diagnosis of a panel of doctors, and is improving every day.
AI is also being linked to robotics; we have seen this with self-driven cars. But the reality is on the West Coast of America. Large articulated trucks have been using this system for years. So, we have here a combination of AI and robotics on a scale never seen before.
The sheer size and scale of what’s possible is incredible. Fox Conn used to pay its workers $2.50 an hour. But it was cheaper to replace the workers with Fox Bots (small AI Robots); a decision to change over to this system resulted in AI Fox Bots replacing 40,000 people in their Chinese factory (Diamansis 2017).
Other predictions are it is estimated that 47 percent of U.S. labor is likely to be replaced by automation (F.com 2017).
Cognitive Engagement
This is where AI technology is used to interact with us humans. Examples are where organizations have a customer interface, which is entirely driven by AI. Intelligent agents are available 24/7 to help customers and provide them with correct information. The medical technology giant Beckton Dickinson in the United States is using lifelike intelligent agent advertiser Amelia to serve as an internal employee in its helpdesk for IT support. SE Bank recently made Amelia available to customers to test its performance and the customer response (Davenport and Ronanki 2018). But, it is also a fact that such agents are not being effectively used by some organizations mainly because of conservatism and misunderstanding of how beneficial this technology can be.
In our personal lives, we seem to have already embraced Apple’s Siri, Amazon’s Alexa, Google Assistant, and Microsoft Cortana, all of which are forms of AI.
AI Today and Case Studies
Today, AI is very advanced; let me quote from Prof M Kosinski to buttress this to stress point.
Computer algorithms, deep learning models are now way too complicated for humans to understand.
AI will impact every facet of our lives for those of us in HR. But for this to happen, we need to make effective changes to our business strategies in order to prepare and manage this paradigm shift adequately. This change is happening now. For HR professionals, there has never been such a big challenge and opportunity. Some assistance to use AI may be sought from the following:
• Google Tenser Flow. It is the AI software that Google uses, and it is a library of information that is available for free. Google is devising plans to run 80 percent of all the world’s AI applications on this platform.
• Tenser Flow is currently being used by the U.S. Army (Summerlad 2018).
• Facebook has open-sourced its neural network libraries.
• Microsoft has open-sourced its computational network tool kit.
• The new-age Quantum Computers is now available, and some companies like D Wave are allowing free access to this.
• In August 2018, Apple became the first trillion dollar company in the United States (Gurman 2018).
• In China, they are currently producing $5 chipsets that give AI enablement (Diamansis 2017). China is leading the world in AI start-ups.
Russian President Vladimir Putin, in a CNN interview, made the following comment:
‘Artificial intelligence is the future, not only of Russia, but of all of mankind,’ Russian President Vladimir Putin said. ‘Whoever becomes the leader in this sphere will become the ruler of the world’ (Gigova 2017).
Algorithms feed AI; this is an entirely different technique that is used for problem solving today. This year, we have seen people solve challenging problems in the field of medicine without having any medical background or experience. This is the uniqueness of algorithms and deep learning.
The following is a quick definition of an algorithm:
An algorithm is a detailed series of instructions for carrying out an operation or solving a problem. In a nontechnical context, we use algorithms in everyday tasks, such as a recipe to bake a cake or a do-it-yourself handbook, as mentioned earlier. (Study.com 2018)
AI and Some Case Studies from IBM Watson
Case 1
Crédit Mutuel
Crédit Mutuel has trained IBM Watson to help its client advisers provide customers with quick and comprehensive information on a whole set of offerings, from car and housing insurance to a range of savings and investment products. “It is impossible for our customer advisors to know all of our 200 products. So we provide them with tools to have the right information for the right client,” said Mathieu Dehestru, Head of Transformation, Marketing and Big Data at Crédit Mutuel insurance. “Watson gives more time to our client advisors, so they have more time for client relationships.”
Thanks to its Watson-powered e-mail analyzer and its four virtual assistants, Crédit Mutuel is enriching interactions between client advisers and customers. Watson has made it possible to find the right answers to problems 60 percent faster. It helps deflect and address 50 percent of the 350,000 daily e-mails received by the bank’s client advisers.
Watson has absorbed over 600,000 pages of documentation, from reports to correspondence. The machine-learning model has been continuously updated to be able to analyze a higher volume of records.
Over 80 percent of Crédit Mutuel employees have adopted Watson for their day-to-day work. Earlier, these employees used to spend 80 percent of their time researching problems and 20 percent fixing it. Now, Watson has reversed this trend.
Case 2
Woodside
Before Watson took charge, Woodside’s engineers spent up to 80 percent of their time trying to uncover possible solutions or hazards, and only 20 percent of their time on the actual engineering work. With Watson, the time spent on researching has been reduced by 75 percent, because Watson enables easy access to decades of wisdom and learning built up by Woodside’s employees.
Korean Air has a year’s worth of historical maintenance records for hundreds of aircraft in its fleet. However, until recently, this vast amount of critical data was virtually unsearchable. This meant that maintenance technicians had to diagnose and fix issues without being able to tap into or interpret implications from valuable past learning and courses of action.
Watson ingested structured and unstructured data from multiple sources, including technical guidelines, nonroutine logs, technician notes, inventory, troubleshooting time and material cost data, and in-flight incident history.
Watson Explorer, using Natural Language Understanding, and advanced content analytics have enabled previously hidden connections that now help maintenance crews to diagnose and solve problems more quickly, with more confidence. Further, if an issue occurs in flight, the cabin crew can report it immediately to ground operations. Watson will access data from similar issues in the past and compare this information against technical guidelines including necessary materials and fixing time. Maintenance technicians fix the issue on the ground and enter their actions into the system to add to Watson’s knowledge.
With the help of Watson, maintenance managers can also identify the trends of issues in each season and can take these insights to the original equipment manufacturers for improvement. Over 200,000 maintenance cases per year are addressed 90 percent faster.
Korean Air needs their over 2,000 maintenance employees to be able to act faster. When Watson delivered actionable insights on the root causes of problems and their solutions, Korean Air shortened its maintenance defect history analysis lead times by 90 percent.
The maintenance employees can now see patterns of defect and failure on equipment quickly so as to take preventive steps in their work. Such preventive measures also allow them to spend more time getting people places on time in their fleet and to work to keep their 25 million passengers happy.
To conclude this chapter, here is an exciting quote from Elon Musk:
‘I am quite close, I am very close, to the cutting edge in A.I. and it scares the hell out of me,’ said Musk. ‘It is capable of vastly more than almost anyone knows and the rate of improvement is exponential’ (Musk 2018).
On the topic of homeworking the following facts cannot be ignored:
77% of those who work remotely at least a few times per month show increased productivity, with 30% doing more work in less time and 24% doing more work in the same period of time according to a survey.
Apollo Technical Engineered Talent Solutions. 6.2021 Working remotely can improve productivity.
Without using spyware or capturing keystrokes, a California-based company has tracked a 47% increase in worker productivity. Based on non-invasive technology that doesn’t grab user passwords, credit card info or other sensitive data, an eye-opening survey shows that smart companies are gaining ground by having workers work from home. What does the data tell us about how employees and team leaders can maximize output during the new normal?
According to survey data compiled from 100 million data points across 30,000 users, here’s how team members are making the most out of their home office:
• The average worker starts work at 8:32 a.m. and ends work at 5:38 p.m
• Tuesday, Wednesday and Thursday are the most productive days, in that order.