Navigating Ethical Challenges in Emerging Technologies

For this assignment I will be looking at the ethical issues raised from emerging technology within my own business sector and two others. My own sector is Data Analysis, for the two other sectors to look at I have chosen Expert Systems and Customer Service. These three choices tie in very well for the department I work in at Bosch, After Sales Service.

Expert Systems

History of Expert Systems

John McCarty, known as the Father of Artificial Intelligence (AI), coined the term “Artificial Intelligence” in his 1955 proposal for the 1956 Dartmouth Conference. McCarthy’s proposal was to look at ways of making a machine reason like a human. A machine that was capable of logical thinking, problem solving and having the ability for self advancement.

Expert systems, although not technically AI, were first introduced 10 years later by Edward Feigenbaum and Joshua Lederberg in 1965. Feigenbaum and Lederberg’s expert system was designed to analyse chemical compounds.

Jump forward to today and expert systems have found commercial use in medical diagnosis, engineering and investing.

But what is an expert system? Britannica defines an expert system as “…a computer program that uses artificial intelligence methods to solve problems within a specialised domain that ordinarily requires human expertise.”[1]

Building an Expert System

Before we can look at the ethics of an expert system, we need to understand how one is created. Depending on sources, an expert system can have either two or five critical components. However, all sources agree that a knowledge base and an inference engine are essential to an expert system. The knowledge base is created by humans who specialise in the area of expertise for the expert system. The inference engineer interprets and evaluates the facts in the knowledge base to provide a user with a result or answer.

Starting with the knowledge base, this is created by a human expert so there is human ethics introduced at this early stage of the process. Although someone can try to remain objective about the knowledge base they are creating, they can unintentionally infer their own experiences into the knowledge base. Similarly, not all experts will agree and one experts response could be different to another’s. This brings us to our first question, can the knowledge in an expert system be reliable or trusted? To answer this we should look at an expert who is trusted on a daily basis, a doctor. Millions of people world wide visit a doctor, provide the doctor with the symptoms of their ailment and wait for the doctor to provide an answer. We accept the doctor is an expert because they have to have a certain amount of knowledge to be in their field. The diagnosis isn’t usually seen as untrue. We see the doctor as a reliable expert and take their word, their diagnosis, as the truth. Doctors do make mistakes, but this rarely sees the doctor from being cast out of their profession. Despite the possibility of inaccuracies, the expert knowledge is taken as truth. It is our own perception that this expert knows what they are talking about that allows us to take this knowledge as truth. If we apply the same logic to an expert system we will trust the source of the system. We are told that the system is expert, we expect it to have the knowledge required to complete the task, we assume that the answers provided will be the truth.

The key back bone to an expert system is the simple rule if this, then that. If what is input into the expert system matches a specific set of criteria, then continue with this specific instruction/step. Multiple ITTT’s can be stringed together to provide answers to difficult questions.

Within the Worcester, Bosch Group Contact Centre there is a “simple” expert system used to repair a customer’s boiler fault by the Contact Centre agent. The agent does not need to have any technical expertise to use this tool, it is a series of yes/no questions that, one by one, eliminates possibilities or provides an answer. The tool prompts the agent with a question to ask the customer, and depending on the customer’s answer the result will be a resolution to the problem or it will direct the agent to the next question. If at the end of the questions the customer’s boiler issue has not been resolved, then the agent is directed to booking an appointment with an engineer to attend and repair the customer’s boiler. This system is used by all agent’s within the Contact Centre on a daily basis, resolving customer boiler issues without ever having to go to a customer’s house. This not only has the customer’s product working within a matter of minutes, but also stops an appointment being booked for an engineer to attend when the problem was something the customer could have resolved their self. This, in turn, frees our engineer support for customers who do need their boiler repaired. Additionally, sending out an engineer to a property when it isn’t a boiler fault costs the business a lot of money.

Ethical Impact

With the example of the Troubleshooting Tool used by the Contact Centre, let us have a look at the ethical impact of the system. There are two main examples that come to my mind, the impact on the customer and the impact on the business.

The impact on the customer can be considerable. The agent in the contact centre is providing a set of instructions to the customer. If the diagnosis is wrong or poor instructions are provided then the customer may be left without a working product. With a boiler providing heating and hot water for most properties within the UK, the consequences of this are exceptionally negative to the customer. The prime time for a customer to call the contact centre is during the colder months where their product will be needed most. A customer without these basic necessities can be left emotional, upset, angry. Having worked with the contact centre I have first hand experience of customers with all these emotions.

The impact to the business may not seem as emotionally based at first, but that can come from a perspective of not knowing the response of a customer to a contact centre agent. Angry customers will usually vent that anger at the first point of contact within the business, the contact centre agent. The level of stress this can put on the agent can be quite considerable and has led to agents becoming emotionally upset from this. Another impact is to the business image. Negative experiences with a business a far more likely to be recorded or passed on than positive ones. Additionally a negative response is more likely to be presented following an emotional impact [2].

Ethical Impacts of Expert Systems

Knowledge maintained within the expert system is only as accurate as the data input by a person. What is the source of the data, how has it been compiled, how is it accessed, how is it presented? What is the impact of incorrectly supplied information? Is the data up-to-date?

Data Analysis

What is Data Analysis

Data analysis is the processing and manipulation of data to observe trends and patterns with a view to providing an informed conclusion or to assist with decision making.

Ethical Impact

Data analysis has, in recent years, come under probably the most potent scrutinisation of an ethical impact. GDPR. The introduction of GDPR has defined how data is acquired, accessed and stored with a person having complete access to, and control of, that data. GDPR was approved in April 2016 with the legislation being enforced in May 2018 [3]. Although it didn’t happen overnight, businesses were “suddenly” presented with a set on instructions of how they could interact with an individuals data.

Working with data in Bosch on a daily basis, our work flows had to change. All data usually came from one large data export which would contain approximately 102 columns of data over 530,000 rows, expanding by roughly 1,300 rows a day. This one data source would be imported, chopped up and sent to different sources within the business which could contain name/address data. With GDPR in effect, a lot of the identifying data had to be stripped out of any business report. The initial impact for us was quite heavy until we had adapted to which columns of data would, routinely, need to be removed. It wasn’t just how we worked that needed to be adjusted though. Raw data for different projects would be stored in their own project folders to allow for easy reference if any queries were to arise. If this data were to contain any personally identifying information, the location of the data and what it stored needed to be recorded. We had to have systems in place to quickly locate any internally maintained personal data, should a customer request a copy of the data we, as a business, maintain on themselves. It has happened, and when address data is used heavily for marketing research, customer contact details for servicing reminders and the processing of appointments as well as any form of contact that customer may have had with us (email or FaceBook for example), the amount of data to retrieve and present can become considerable. Especially if the customer has been with us for many years, some people have been customers since 1984!

The one thing didn’t change for us though was the acquisition of the data, Bosch has always been ethical with how data is acquired. Any information provided to the business by a homeowner was always via direct communication with that customer, via phone, email, FaceBook or Twitter.

The introduction of GDPR though can only be seen as a positive thing. I think everyone can agree that having control over the data a business holds on you and how that data is acquired, as well as having the option to have your data removed from a system, is only a positive step forward in a world where nearly everything you do now can be electronically recorded in some manner.

Customer Service

There are three areas to consider for ethical impact when customer services are considered. The business, the customer, the Customer Service Agent (CSA). Whilst the CSA works for the business, and they can be considered the same entity, I have chosen to separate them as a CSA’s interactions are with a customer directly, whilst the business, as a whole, can impacted differently.

A CSA has a lot of responsibility with their interaction with customers. They are usually the first, and sometimes only, point of contact between a business and an end user. How they conduct themselves during a customer interaction will reflect on the business and can have a significant impact on how the customer will interact with the business in the future.

A CSA will have to ask security questions to establish they are speaking with someone who will have access to the account data. They have to ensure any data they acquire during their contact with the customer is recorded correctly and accurately. The customer can be emotional or angry during their interaction, with the CSA having to respond to this customer state responsibly. All this is during one interaction with a customer, the CSA will have scores of interactions in one working day within a five-day week. With each interaction though, the CSA will remain professional and courteous. With the number of ways a customer can now contact a business, how a CSA retrieves and processes necessary data can vary. Commonly, now, social media is becoming the norm. A negative social media response can quickly be viewed and shared, with a single post, from either the customer or business, becoming “viral”. This has been seen in recent times with EA, the games publisher, receiving a Guinness World Record for the most downvoted comment ever in Reddit, following a response about the release of their game Star Wars Battlefront II. This Reddit comment was a response by EA to other comments [4]. Any information published online by a customer to a forum or social media site can be considered to be in the public domain, and therefore publicly available. Once that data is recorded by a business though, via a complaint or with a view to obtain further information, the data falls under the guidelines of GDPR with the business having to ensure that all guidelines of GDPR are maintained. As the data is publicly available, the acquisition of this data is already covered.

Conclusion

In conclusion, the exploration of ethical issues in emerging technologies across the business sectors of Data Analysis, Expert Systems, and Customer Service reveals the complex landscape of human-machine interactions. From the historical development of expert systems to the ethical impacts on customers and businesses in customer service, and the paradigm shift brought about by GDPR in data analysis, it is evident that ethical considerations play a pivotal role in shaping the responsible use of technology. As businesses navigate the integration of these technologies, a proactive approach to ethical decision-making becomes crucial to uphold trust, ensure customer satisfaction, and foster positive interactions in an ever-evolving digital era.