Automation in the digital world: the loss of human cognition.
In 2017, the Office for National Statistics estimated that about 7.4% of jobs in England were at high risk of being automated in the future, with an increasing number at low and medium risk. Jobs at a higher risk are not limited to those that do not directly interact with people like machine operators. Customer service, recruitment and store assistants are all at risk of being automated. This begs the question: can technology like AI truly replace real human cognition during complex human interaction? Looking at a few examples of this digitalisation of the human brain, we can get a clearer picture of what it brings and what is sacrifices.
Chatbots
Chatbots are computer programs, most often embedded in websites, that facilitate a conversation between a human and the computer. The value of these types of programs to businesses are expected to grow massively over the coming years with the market value of chatbots in banking, financial services and insurance (BFSI) expected to grow from $586 million in 2019 to $6.83 billion in 2030. One type of chatbot are keyword recognition chatbots (common in online customer service situations) which retrieve information by matching that information to individual words it recognises entered by the human. Another type are contextual chatbots which learn about the human users and you converse with it. Previous messages are stored in memory to achieve an end goal like ordering food from your local takeaway.
When used well, chatbots can be effective ways of retaining sufficient customer service levels in the absence of genuine human intelligence. It has the added benefit of being able to work 24/7 and serve multiple customers simultaneously; preventing the dreaded waits in queues to speak to a customer service representative. An example of an effective chatbot is that used by Emirates. The chatbot embedded in their app allows users to message about potential holiday ideas. The chatbot will attempt to recognise keywords based on locations, activities or weather described by the user and match the response to holidays destinations they fly to. With a click on the bot’s response, the user will be redirected to a booking page for the holiday package suggested by Emirates.
So.. Emirates benefits by reducing expenditure on labour for customer service or travel agents, and the consumer benefits by avoiding giant queues without having to look further than the screen in their hand. Chatbots are always a win win then? Unfortunately not.
Chatbots can be effective but some chatbots can be more effective than others — as George Orwell once wrote. ‘TOBi’, the Vodafone chatbot, is less effective than the previously described one from Emirates. From my experience, it struggles to understand questions or instructions and even if it does, the assistance it provides is very limited. For example if you say you would like to buy a phone it provides a very general link to the webpage where you can buy phones, which is easy to navigate to on your own no matter what your technological literacy. Chatbots can also have further reaching implications other than just being ineffective. Some AI chatbots have been compromised with damaging consequences. Microsoft’s ‘Tay’ was designed to be trained by public tweets but within a day learned to become racist. If chatbots are often unnecessary and in some cases learn harmful traits then their existence can be questioned. Is their usefulness diminishing giving the increasing technological ability of society?
AI and Recruitment
With the ever growing ability of artificial intelligence, there are increasing ways in which it is implemented and often in an attempt to replace humans for streamlining and efficiency. It may be unsettling to hear that AI is even deemed capable of determining whether you are worth hiring. AI is often used to check over CVs, run psychometric and numerical tests and even interview you. These methods are most commonly invested in by enterprises who will have vast numbers of applicants each year. In fact, the AI recruitment industry is said to be worth over £350 million!
AI recruitment has benefits both on employer and employee side. Automated recruitment saves 2 days per week of a recruiter’s time on average. It also reduces the average time taken for an applicant to apply from 17 minutes to just 3 minutes! A business can invest these saved resources elsewhere making them more efficient while making their job roles more accessible to budding applicants. As a student preparing to apply to graduate schemes, come the beginning of my final year, the ability to save time on applications that I could be spending on my university studies is very appealing.
There are of course however some more concerning implications of implementing AI in such a way. For example if the software is too simplistic, the employer could be filtering out talent at a very early stage that a human may have been better at noticing. Another more serious issue is that the software used to analyse applications is designed by real people (for now at least) and is therefore open to having prejudices built into the code. Some people who are from ethnic minorities or ‘working class’ backgrounds feel the algorithms “aren’t designed by people like me, and they aren’t designed to pick people like me”. The EU and US are however starting to write legislation to monitor any algorithmic bias which, it could be argued, was not possible when the role was done by humans. It could also be argued that with all applicants taking the same interview judged by the same ‘interviewer’ then variations in bias are irradicated. With the help of written law it may be possible to ensure that the benefits of this time and resource-saving tech can be free from bias.
Conclusion
Whilst weighing up the benefits and drawbacks of automating human roles is important, said automation does almost seem inevitable. As seen with use of AI in recruitment, the benefits seem to be on both sides of the coin. There are bound to be teething problems with such a drastic change, some of them more sinister than others (racist robots and all), but we shouldn’t let these dissuade us from the natural progression of our society.
Reflection
Having been digital enthusiast as long as I can remember, I chose Digital Society because I knew it was going to be something I found interesting and would enjoy. I therefore did not expect to learn as much as I have. My enthusiasm for the digitalisation of everything and anything perhaps blinded me to the numerous implications of what living in a digital society truly means. That is not to say this course made me more resistant to this new way of living, but more street smart (..or web smart).
For years I have been shutting anyone down who criticises social media because it appeared to me that it was just something they didn’t want to understand. However, social media can be used as a weapon against democracy which can be seen in the Brexit vote with Cambridge Analytica and also the 2016 US presidential election with the rhetoric of “fake news”. Social media’s attempt at free speech has been threatened in this way and in my opinion has disrupted the balance of free speech rather than enhanced it. Whether Elon Musk’s purchase of twitter is beneficial or harmful to free speech is up to personal interpretation. This realisation of the harmful impact a digital society can have was aided by week 6 topic “The individual, identity & ethics”. I found the discussion about whether we can apply the same rules from our offline lives to our online lives fascinating and it showed me how different these lives are.
I found the assessment style of this course very interesting and also refreshing. Having been taking a very essay heavy semester, it was enjoyable to write in a more relaxed blog style for this unit. With the digitalisation of society, being able to write interesting and attention-grabbing content for the internet is a crucial skill so I am glad to have got some very valuable experience through this module. This skill is likely to be important for whatever job I do in the future to communicate with clients, customers or colleagues online. Writing for an online audience is definitely something I have found difficult while having to write academically with in text citations, bibliographies etc for most of my university life so far.
I also found the style of how the course was taught enjoyable. It is strange to learn so independently with no in person lectures to meet fellow course mates, but this has encouraged me to become more focussed when learning independently at home. Learning in this way also made the small snippets of interaction in response boxes etc more meaningful and also significant of the digital society topic we were all studying electronically.
Overall this course has challenged my gung-ho approach to the digital world by making me more aware of its implications while also developing my writing and academic practice. This will benefit me beyond my university degree and into my professional career.
