To AI or not to AI

 

AI Hype

NVIDIA’s market cap just hit $4.5T USD — it’s the company powering the AI revolution and the growth in NVIDIA’s value has no basis on its current company value (revenue or profits) — it’s pure speculation based on the hype around the potential of AI.

If you spend a decent amount of time on social media (or LinkedIn, as I do) you can’t avoid AI related content, from the best ChatGPT prompts to the best AI assistants, to “look at this great profile pic I made of myself which is me, but also not me” — AI is everywhere.

So what is all the hype about? What makes AI the next great thing? Why are people willing to gamble on AI related companies? Why are companies looking to lay off workers because of AI? There are a lot of questions and not many solid answers

As a leader in business operations and processes I’ve had the opportunity to play around with a number of AI platforms over the last few months. I’ve built AI bots, assistants, and integrated AI workflows into some of our processes — but I always came back to the same thought — AI is just automation — smarter automation, but automation nonetheless.

AI = Automation, Just Smarter (and Riskier)

My role in most organisations is to understand the processes and operations that support the delivery of a product or service, and to identify opportunities to increase efficiency or reduce costs while maintaining or improving the quality of the product or service.

In my experience, automation is one of the most effective tools to achieving those outcomes. Automation tools empower users to do more with less and it supports organisations to scale efficiently — keeping costs and overheads low relative to the growth in the organisation. When looking at the characteristics of automation and applying it to AI, it’s hard to ignore the similarities:

• Both rely on inputs → logic → outputs.
• Both need to have set rules, data, and governance.
• Both are great at replacing repetitive human effort

The only difference is that AI can learn and adapt, which means when it goes wrong, it can do so spectacularly.

AI Didn’t Rewrite the Rules of Automation — It Just Raised the Stakes

Standard automation is rule based, ie:

• It requires explicit rules such as if X, then Y.
• It’s fixed, and will only do what it is told,
• It’s transparent and explainable
• Practical use through process automation, workflows and Robotic Process Automation (RPA)

AI is pattern based, ie:

• AI has implicit rules which it learns from data
• It’s flexible and can adjust based on patterns
• There is very little transparency (the black box), although, some LLM’s allow you to see the thinking/logic behind the answers
• Practical uses include prediction, generation or interpretation

And more inportantly, unlike traditional automation, AI can introduce bias, hallucinate, or make opaque decisions — and it often does so in ways that aren’t obvious until things break. That’s why the foundational principles of automation matter now more than ever.

The Golden Rules of Automation Still Apply

MY NUMBER ONE RULE IS: Dont use automation to fix bad processes.

Make sure you follow the 4 golden rules of automation to see the greatest impact from your use of AI tools:

1. Fix (or understand) the process first. Automate stability, not chaos.
2. Start small and iterate. Big-bang automation rarely works — and with AI, it can create black-box failures.
3. Validate and monitor constantly. AI needs ongoing oversight, not set-and-forget.
4. Governance matters more than ever. Because when AI fails, it fails fast and at scale.

Governance as the Differentiator

I spent a few years in government and public health and governance is everywhere — it’s one of the main reasons why things move so slowly there — but there is a good reason for the risk aversion. Providing a public service — especially in health — requires robust policies and processes to ensure that people receive the best possible care, are safe and looked while under care and continue to trust the public health system.

When it comes to AI, governance is also important to ensure that people receive the outcomes they are looking for, all reasonable risks have been mitigated and people can continue to trust the processes that are in place to deliver the product or service.

AI Automation principles for Real Value

My view is that AI is worth the hype, however, it is still an automation tool that we have not quite figured out yet. If used correctly, AI will be one of the greatest tools we have at our disposal to improve the lives of people across the globe, the environment we live in, and provide a better future for the next generation.

In business, I believe it will have a greater impact than the spreadsheet — there are too many possibilities to think of, but if it is used wisely and the right governance is in place it will:

• increase the capacity of people in their roles (not to mention how founders and entrepreneurs can use these tools to support them through some of the most challenging times of a startups),
• be able to process and analyse large amounts of data effectively which will enhance decision making, and
• level the playing field for small and medium businesses competing against large (and global) companies

For operational leaders, the message is clear: don’t chase the AI hype. Focus on process discipline, good governance, and meaningful outcomes. AI is a lever, not a ssilver bullet — and its value is unlocked by the maturity of the systems and people around it.

When delivering talks on automation, I always used the analogy of a building site — you can use a shovel and basic tools to build a house, or you can use diggers, cranes and other technology to speed up the process. If there are unseen hazards like water pipes underground or there are no safety policies, the damage that can be caused by a shovel is a lot less than what can be done by a digger or crane — more powerful tools needs greater oversight and planning to reduce the risk of damage.