Automation is everywhere. But nobody wants to be put out of a job by a robot. The idea that the workplace skills that any of us have developed over the years could be digitally encoded and then delivered by a hardware or software robot is always a little disquieting.
But perhaps we should have understood these trends a long time ago. Automotive manufacturing plants in the UK have been using robots since the 1970s - hence the famous ‘hand-built by Roberts’ sketch on Not The Nine O’Clock News back in the day. Moving into the 1980s, Japan has been levying union dues on deployed robots for some time now. This stuff is real.
The modern era’s software robots, chatbots and Robotic Process Automation (RPA) automations were on the table a long time ago. Today, it’s a question of which job will get automated next and how well humans will embrace autonomous virtual workers and look to move to higher-value tasks and roles not yet within robotic reach, so-to-speak.
Ripe for automation?
At this risk of touching a tender nerve, the practice of press public relations has (arguably) a number of operational practices that may soon be automated-out of the purview of the keen-eyed PR executive.
“Automation is great for building manual pattern-heavy reports outside the chaos of media relations but, when it comes to human creativity and relationship building, automation sometimes falls flat. It’s hugely useful for pattern recognition, but doesn’t perform in environments with extreme unpredictability where the nuances of human behaviour are at the fore,” said Jack Buckley, technology PR manager and member of the admin team on UK social media press-PR channel TechJPR.
Buckley highlights a key truth i.e. there are products out there that claim to automate PR end-to-end, but the irony is that the people who are really good at PR aren’t the ones building automation tools. He argues that there is a skills gap and disconnect here from the outset. If PR (or some other human interfacing function) was as simple as running a Natural Language Processing (NLP) analysis on media articles, it would have been done already.
“The journalists we work with receive thousands of emails a day. The answer to cutting through that noise isn’t sending more emails faster – it’s building mutual trust, focusing on quality, and sending better pitches,” added Buckley.
Looking more broadly, we know that Customer Relationship Management (CRM) already has a high level of AI-driven intelligence and software bots already at work. This union of people of tech has been mostly harmonious - spoiler alert: you may have noticed Salesforce is relatively successful - so we will have to get used to more bots very soon indeed.
A definable level of repeatable functions
Whether public relations or some other admin-heavy job function starts to be increasingly automated is currently open for debate. It seems clear that essentially office-based roles are the most suitable candidates for automation, especially ones where there is a definable level of repeatable functions and operations i.e. the stuff people hate and computers love.
Perhaps our next most likely to flourish AI space is recruitment. Sense thinks so. The company is an AI-driven talent engagement specialist for recruitment functions. Now working with MLOps platform company Iguazio, Sense is using automation and AI to speed up the recruitment process. Immediately distancing itself from any notion of cold robotics, Sense is delivering what it promises is a hyper-personalised candidate experience.
Sense provides the Sense AI Chatbot. This is an automated recruiting assistant that can engage with candidates 24/7, responding to their queries in real-time even when human recruiters are offline. It engages with candidates across SMS, mobile and the web to match people to jobs, schedule interviews, and handle ‘intelligent communication’ interactions, which in this case means FAQs.
Conversational voice & text AI
The chatbot pairs conversational AI with automated communication and engagement workflows. Sense has a large team of data scientists and machine learning engineers with deep expertise in conversational AI – both voice and text. The team’s challenge was building the complex natural language processing (NLP) serving pipeline, with custom model ensembles, to track question-to-question context and enable sentiment tracking.
“Off-the-shelf chatbot solutions were not sufficient. For this complex MLOps project, we needed that powerful mix of high-performance real-time serving graphs, Nvidia [Graphical Processing Unit - GPU] accelerated computing, plus the ability to scale up and down based on load,” said Alex Rosen, co-founder and head of product at Sense. “Iguazio was the best MLOps platform that provided all of these capabilities.”
Sense chose Iguazio to build a robust, automated, real-time NLP pipeline. The solution uses the connectivity of Iguazio’s MLOps platform and built-in feature store with the Snowflake Data Cloud to speed up feature engineering, deep integration on Amazon Web Services (AWS) to manage cloud consumption costs, and smart scheduling capabilities for Nvidia accelerated computing to manage GPU usage in an efficient and scalable way.
Sense selected Iguazio through AWS Marketplace, which allowed all parties to accelerate and automate the procurement process. The Iguazio platform is deployed in AWS GovCloud, which gives government customers and their partners the flexibility to architect secure cloud solutions. Corporate box-ticking notwithstanding, these actions do at least underline the need for strict compliance as we start to roll out more software bots.
Our bot-boosted future
Over and above the work Sense and Iguazio have carried out here, what trends can we take away from the technology being implemented here? There certainly seems to be a level of hopeful trust in the ability of AI platforms using chat automation to actually carry out intelligent conversations.
If an FAQ is as intelligent as you need, that’s great; but ask anyone who’s been on an airline chatbot discussion about their experience in the post-pandemic scramble that we have witnessed across global airports in 2022, feedback is likely to be less than positive in many cases.
Let’s remember the first three rules of economics here to guide our opinions in this space. There’s no such thing as a free lunch (so for every bot there is a payoff somewhere); everything is cyclical (so let’s temper our enthusiasm where we can); and thirdly, the safest job in business is undertaking (demand is never in question)... and the bots won’t be taking on that role anytime soon.