AI has cracked the knowledge workplace

Managing for Society
The Manila Times
January 3, 2023

In 2016, Klaus Schwab, Founder and Chairman of the World Economic Forum, declared that the Fourth Industrial Revolution would impact our lives completely. Every industry would be digitalized, disintermediated, and new business models must be defined.  Moreover, because of the danger of job losses, there would be a need to reskill and upskill people fast. Schwab was warning about how fast information, biological, and nano technologies were advancing and encroaching into more and more aspects of our lives. Analysts predicted these developments would impact half of all jobs within five years.

Five years have passed, and analysts grossly underestimated the speed at which technological advances will affect how human beings work. With the improvements in smartphones, high-speed networks, and the Internet of Things, we increased our use of various artificial intelligence (AI) tools to enhance our work lives and beyond without even thinking about it.

In the morning, we ask our personal digital assistants how to cook Eggs Benedict for breakfast. Our refrigerators tell us what we need to buy at the supermarket. Our vacuum robots clean our floors when we leave the house. We ask our smartphones for the fastest route to drive to work.  At work, software corrects the grammar in our reports, presentations, and emails. Software automatically analyzes and generates recommendations based on the steady stream of data our company collects. When we get home, video streaming services recommend movies we will likely enjoy. And when we go to bed, our smartwatches monitor the depth and quality of our sleep. I could go on.

But 2022 will be remembered as the year AI significantly cracked the knowledge workplace. AI tools are now freely available for two tasks we typically reserve for humans: the creative generation of technical communications and the production of presentations.

In September, Tome opened its AI-powered storytelling generator to the public, enabling anyone to automatically generate a draft slide presentation on a specified topic with text and accompanying images. I asked Tome to prepare a presentation on “Promoting ethics in the workplace.” In a minute, it generated an eight-slide presentation titled “Ethics in the workplace: A new frontier.” It is a draft presentation, to be sure, but the text and images are an excellent start which would have taken me at least an hour to conceptualize and prepare in slide format.

By refining and customizing Tome’s output, I could give a presentation within two days’ notice for topics I am familiar with. Since I usually allow two weeks of preparation when accepting speaking engagements under similar circumstances, Tome could increase my consulting productivity tremendously.

In November, OpenAI released Chat GPT, a freely available natural language AI model that can interact conversationally and creatively on virtually any topic. I wrote my first column in 1994 (also on technology – the World Wide Web). I’ve been writing columns since, and it typically takes me at least a week to research and write even a short piece. I asked Chat GPT to write a column on “Motivating the Gen Z worker,” and it typed out 600 words in just three clicks in flawless English and credible prose. Is it original writing? I copy-pasted sentences from the column into Google, and nothing identical came up. Allowing time for me to validate and refine the column to my style would still give me at least a ten-fold productivity increase!

I next told Chat GPT: “Write a polite letter refusing a customer a refund for a product which was likely damaged by the customer.” The letter was typed out in 15 seconds, and I thought it was perfectly human-like and professional.

While the AI examples above are jaw-dropping, they raise some issues. Firstly, the user could be questioned on originality and integrity if AI tools are not disclosed. This could lead to a flood of plagiarism – dubbed “AIgiarism” -- charges in the case of student use, for example. Furthermore, the creativity and learning of AI are nothing like that of humans. AI tools produce answers by mathematically detecting statistical patterns in large volumes of digitized human inputs, mainly from the internet, and then putting these patterns together in new and seemingly human ways upon request.

Hence, these tools are always limited by the data and mathematical algorithms fed into them. Finally, these tools have been shown to often produce errors that would not be noticed by naïve users. Hence, blindly relying on them could be disastrous.

Clearly, the technology-enabled future Schwab predicted has arrived as far as office knowledge work is concerned. All workers must understand how to use the increasingly available AI tools effectively and ethically, mindful of their limitations. Used properly, they have the potential to free us from the routine and data-intensive work that computers do so well. We can then focus on innovatively interacting with our customers and stakeholders to genuinely improve their lives.

Dr. Benito Teehankee is the Jose E. Cuisia Professor of Business Ethics at De La Salle University. Email: benito.teehankee@dlsu.edu.ph