Artificial Intelligence: The Evolution of the Human Workplace

Zeeshan Javed, Head of Media Performance and Analytics, Foxtel Group

Zeeshan Javed, Head of Media Performance and Analytics, Foxtel Group

In the past 70 years, few moments in the technological evolution of humankind have been more prominent catalysts for industry-wide change than the rise of artificial intelligence (AI). Be it the birth of the internet, the release of Apple’s first iPhone, the rise of the social media behemoths, the great remote working migration during the pandemic or AI and LLM, each in their own way transformed both business and personal capabilities overnight.

However, AI poses a uniquely critical question for workers everywhere: ‘Is AI going to replace human involvement in the workplace?’ To answer this, let us explore the evolution of AI into what it has become today.

The Early AI Revolution: Alan Turing, Neural Networks, and the Robotics Revolution

In the 1950s, Alan Turing proposed the Turing test to see if machines could provide responses to questions that were indistinguishable from human responses. The premise was simple: Humans have the ability to consider context within both verbal and non-verbal responses, and a machine would have to perform the same neural calculations as that of the synapse-connected human brain to mimic the behaviours.

Once Pandora's box opened, the evolution of AI sped up with the development of AI languages (LISP), machine learning models, neural networks, robotics and much more. Capability grew on the hardware side with the growth of quantum and cloud-based computing, which, through its calculative capabilities, put AI’s end goal within reach, and it was a matter of time before software and core functions caught up. All of this paved the way for AI as we know it today.

Chat-GPT is Born

When we talk about fundamental shifts in AI capability, Chat-GPT ( Chat Generative Pre-Trained transformer) has been a pivotal evolution in this AI revolution, giving birth to what will rapidly become a trillion-dollar industry that permeates all corners of society. By fundamentally being able to converse with its questioner and adapt its responses across 175 billion different parameters, it changed the type of conversations a human could have with a machine. Its successor, GPT-4, has exponentially expanded that capability to parameters rumoured to be in the trillions, an ability to assess both text and images and perform countless permutations in seconds. All this has given credence to an inevitability that it will outgrow any technological evolution that came before it in a fraction of the time.

AI is the Great Equaliser

The open-source nature of LLMs has enhanced fundamental capability within multiple industries and has evened out the playing field, where even new start-ups and SMBs can compete with established behemoths.

The thousands of companies born via API add-ons to the transformer engines have allowed a new range of functionality that used to be left to human-based consulting services. Resumes, videos, translations, call centre responses, and much more can be managed in seconds via prompts. On a personal level, as a data and digital professional, I can see how these models now put access to what were previously data science team-led capabilities like marketing mix models directly into my hands, reducing reliance on external consultancies.

Changing of the Guard: Prompts Replace Coding, Which Makes Data Accessible

Previously, knowledge of SQL, Python and other programming languages was critical to extracting insights out of large data sets, especially your cloud-based environments and clean rooms. The evolution of AI has changed this with prompt-based engineering. This means that even junior analysts can now extract data without programming language capability, and more than coding, prompting is the new skill set to have. This fundamental change opens up data extraction and insights procurement to a whole new generation of data scientists without requiring the mandatory technical skillsets of yesteryear.

Not All Labour Is the Same

Not all industries will be impacted equally, but in some cases, AI can replace the menial tasks that junior analysts used to do. This includes basic copywriting, data summarisation, presentation development, customer aftercare services and much more. Now, does this mean that human resources is no longer required? The answer depends on the core mentality of the organisation employing it and the customers at the heart of their operations. You see, human beings are not a fully rational species, and as such, an AI platform built on reason and logic cannot fully address the cognitive imbalances in human responses. Jobs such as business strategists that walk the line between logic and instinct will not be replaced in the short term, but repetitive tasks that distract resources, such as scheduling and data aggregation, can be.

The open-source nature of LLMs has enhanced fundamental capability within multiple industries and has evened out the playing field, where even new start-ups and SMBs can compete with established behemoths.

All Great Technological Advancements Carry Risk

While AI has multiple use cases, it still mirrors the birth of many impressive technological evolutions like the internet, where, at the beginning, guardrails and ethical guidelines did not exist. This led to a bombardment of poor content, services, and the darker side of internet evolution: Dark web and malware. For me personally, the biggest hesitation to the broad adoption of LLMs and AI practices is the high-risk exposure that an ungoverned AI poses to an organisation’s proprietary equity. SMBs may be more willing to take risks, but larger organisations will undertake comprehensive checks and balances before they hand over the keys to the kingdom.

The Future of Labour in the Workforce and AI as an Enabler, Not a Replacement

It stands to reason that AI was created to simplify and minimise certain tasks that were ill-suited to human beings to begin with. Complex calculations that would take a machine milliseconds could take a human the equivalent of multiple lifetimes. In this way, AI is doing exactly what it is intended to be: an agent of change, allowing for deployments to be done faster. In summation, AI has a multitude of roles in the future workplace. Some industries will become risk-averse early adopters and, move quickly and ignore the weight of data governance. However, in a great many industries, only once the risk mitigation and relentless testing have been done, AI has the potential to be an enabler, allowing for greater workforce utilisation on work that matters.