Moments that Matter - Part 2
Introduction
Matthew 22:39 [ESV] “You shall love your neighbour as yourself.”
Over the last 2-3 years we have seen the emergence of Large Language Models (LLM) capable of generating text given access to input information. Agentic AI is on the rise automating a wide range of human/machine interactions. In this article I want to address some interactions which historically have been human-to-human but are beginning to become more autonomous machine-to-human interactions. I wonder whether all these applications are justified, helpful or appropriate – and whether we are living up to Jesus’ challenge of loving our neighbour as ourselves when we embrace them
AI Filtering Job Applications
Using AI tools to filter job applications is a common practice. At some levels, it feels like a necessity given that tools like LinkedIn can enable simple applications which result in many 100’s of candidate resume’s landing on the hiring managers’ desks. Yet despite claims from the software vendors that these tools are accurate, reliable, effective ways to create candidate shortlists, there are documented cases of biased filtering which have caused these tools to be taken out of service, at least temporarily. In my own experience I know that extremely well-qualified candidates are being rejected within minutes of submitting applications by these tools. Given that ever more candidates are using AI tools to write their resumes we are in something of an ‘arms race’ in the jobs market. AI is assessing AI. Whatever else is happening, this is not really a fair assessment of candidates. I think the current approach to recruitment is lazy. It lacks humanity and is unlikely to raise the quality of the workforce despite the grandiose claims of recruitment software vendors. It is likely to amplify biases – whether in the form of the target job descriptions or candidate profiles. We must do better.
Using Generative AI in personal communications
Take the time to talk
Face-to-face if possible
Generative AI offers some impressive capabilities to generate (usually) meaningful text and save time generating volumes of content which can be tailored effectively to specific needs. It is tempting to use this capability in all sorts of applications including things like:
Performance reviews
Student/School reports
General feedback and assessments
As managers, teachers or tutors, I believe part of our responsibility is to supervise or teach those we are responsible for as fellow humans. Sometimes, giving feedback is difficult – particularly when some of what we feel we should say can be seen as critical. Resorting to a machine to provide that commentary, risks our shrugging off that responsibility and passing accountability for difficult messages on the AI.
Just because we pass the inputs to a machine does not guarantee accuracy or consistency and, general-purpose AIs have no ‘skill’ in empathy, or in delivering sensitive messages. These moments of feedback are moments that matter in the lives of the recipients. I would suggest our responsibility is to provide holistic feedback, in person, human-to-human, rather than passing data to a machine and letting it generate an automated impersonal response.
Customer Service
Just because we can, does not mean we should.
Many customer-facing businesses have removed or hidden call-centre numbers and replaced many of their customer-service agents with chatbots. How many times, when an order has been mis-delivered, gone missing, or the goods supplied arrived damaged have you desperately wanted to talk to someone to resolve the problem, when the automated agent cannot exactly decode your problem and respond appropriately. At this point, chatbot implementations are not good enough completely to replace human beings. In the race for profit, humans have been let go, to allow poor digital facsimiles take their place, leaving a significant percentage of customers frustrated, and feeling they have little or no recourse.
If we then move into the domain of financial difficulties – where people are struggling to meet financial obligations. Hard-wired digital engines may make consistent decisions here – but they rarely comprehend the specific circumstances faced by a struggling family, dealing with a seriously ill family member, or perhaps they have lost work because a job has been lost to automation, and alternative work is not forthcoming. We are facing radical social change in western society as AI Agents ‘move in’. Leaving the AIs to deal with the consequences is perhaps the ultimate abdication of human responsibility.
Conclusion
Technology innovation is moving at an incredible pace. It is possible that rate of change may slow, but there are no obvious signs of a slow-down at present. Let’s look for and embrace benefit where it can be found – but let us be as enthusiastic about protecting our humanity as well.
In our dealings with our family members, our colleagues, our students, our pupils, our acquaintances and our friends; let us rely on simple in-person human-to-human interaction wherever we can. It will sometimes be messy, but ultimately this allows us to be fully human in relational world. Anything less is a choice to live as something less: perhaps sub-human rather than the superhuman we imagine we might become with our technology super-powers. I hope we will seek to treat our fellow humans just as we wish to be treated ourselves.
Photo by Jessica Da Rosa on Unsplash