Alex Croucher, Founder and Intelligent Automation Director at VKY Intelligent Automation
FAKE news has reached a fever pitch in recent years. From politics to urban legends, it can at times be a challenge at times to figure out the fact from fiction. The concept of myths goes as far back as Greek mythology.
In the technology world, we have seen a flood of fake news created concerning artificial intelligence (AI). Here we will disprove the top four that are on the tip of many people’s tongues.
Myth One – AI will steal everyone’s jobs
After the emergence of Chat GPT, this was perhaps one of the most common concerns we heard from the market. However, this opinion is starting to wane.
In a recent report ‘What Workers Want 2023: Working with AI’, HAYS surveyed over 9,000 professionals in the UK and found that over half of respondents felt positively about AI and believed it should be embraced in the workplace.
The theme of change is nothing new when it comes to how we work. When spreadsheets were first developed, bookkeepers were concerned they’d be out of a job and when Henry Ford developed the factory line, workers thought the machines would make them redundant. In reality, these new developments supported people in their roles rather than replaced them. Employees simply saw a change in the type of tasks they carried out – and in many cases, their jobs were made a little easier.
The most successful businesses we see just now are the ones that use AI to assist their workforce. Combining people skills with AI is powerful, and arguably crucial if organisations want to stay competitive, especially in areas like customer service.
In the same HAYS report mentioned above, 38% of the respondents said customer service and contact centres would be the top profession to be impacted by AI.
For example, organisations that have incorporated AI into their customer service models are seeing higher customer and employee satisfaction rates as well as reduced operational costs. Assisted by AI, employees can focus on more meaningful interactions with customers and let AI worry about data entry, efficient processes, and 24/7 omnichannel customer contact.
So, rather than mass redundancies, the rise in the use of AI is likely to see job roles evolve and adapt.
Myth Two – AI can do the job of humans, better
Very closely linked to our first myth, is the notion that AI can not only do the job of a human but do it better. This is not true.
There are varying degrees and types of AI, but no AI adopted into business will be able to develop customer and partner relationships in the same way humans can. In terms of creativity and ideas, these also need a human starting point. For example, generative AI is only as good as the instructions it is given by a human.
Building on the customer service example, although AI will impact this sector significantly, the AI solutions we provide to customer care and contact teams couldn’t work without people. AI cannot deal with complex customer inquiries or needs. Nor can it empathise with the customer in circumstances that are out of the ordinary.
What AI does do very well, however, is set the scene for people; doing all the data entry and analysis in the background so the human can complete their work in less time and with fewer obstacles.
Rather than see the threat, we need to acknowledge that AI and humans are better and more efficient when they work together.
Myth Three – Adopting AI means losing control
It almost goes without saying that for anything – not just AI – if left without any guardrails, mayhem ensues.
However, when it comes to AI, discussions are taking place to bring a level of governance to the technology. While it might take some time for governments to put in place robust national /international regulations, it is possible for organisations to establish internal frameworks to maintain control over data and limit any privacy fallouts when using AI.
To get to that point though requires a level of human-led training. On its own, AI cannot understand anything about the specifics of an organisation let alone help customers as it cannot automatically connect to other internal systems. This kind of training is important as when it comes to generative AI, it can be prone to “hallucination” where it can make up incorrect facts.
The good news is that we are starting to see enterprise level generative AI coming to the market – most recently from OpenAI which will ensure that sensitive company data is not used as part of its wider machine learning to evolve the platform. When consumed through Microsoft Azure OpenAI Service, it’s possible to ensure data residency in the UK, or provide private non-internet based connectivity into enterprise systems. This means organisations can take advantage of AI, without sacrificing the security of their sensitive data. Use of this variety of generative AI will limit control loss and align with industry data regulations and governance models.
Myth Four – AI is only something large businesses can afford
You don’t have to do a big bang with AI. The ubiquity of this technology means that we have moved beyond the days of multimillion-pound investments into systems like IBM’s Watson and can use readily available AI tools that in some cases can be free to enhance what they do.
Many of our customers adopt and implement AI in phases. This can not only help with cost but also internal adoption, ensuring there is time to upskill teams and guarantee that a culture is created where its adoption will be a success. It is our experience that it is quite often the simplest of implementations running in the background, that no one sees or notices, are the ones that drive the most ROI.
This is especially true of conversational AI which, at its core, is about understanding customer intent and interacting using natural language. It could be used to replace the traditional, and often frustrating, “press option one for…” situation and allows callers to simply explain what they want before being routed to the most appropriate outcome. Things can even be taken a stage further, with identification and verification being completed by an AI virtual agent, resulting in a quicker response for the customer and allowing the human advisor to spend their time managing the customer relationship and solving more complex enquiries.
If history is anything to go by, we will all quickly learn the myths that currently exist around AI will completely disappear as our knowledge and experience of the technology increases.
Rather than seeing the threat at large, we must consider the wider benefits that will be seen for not only organisations but also individuals over the long term. I have no doubt we will get there, but it will be in the seeing that we all end up believing in AI’s potential.