In the previous article, "ChatGPT (OpenAI)", we discussed lessons from the internet revolution, and finished by asking, should we even use AI at all in our businesses?
Decisions outside of our circle of control
Asking such a question implies there’s still such a decision to be made. But there isn’t.
Whether or not to use AI is no longer up for debate.
AI has already been chosen for us. By the market, by your suppliers, and by the advancement of modern technology.
There’s nothing that you can do about that. It’s here.
AI features are already built-in
Every business system you rely on — Microsoft 365, Salesforce, SAP, Adobe, HubSpot — now includes AI functions by default.
AI for business currently falls into 2 broad categories - generalist AI chatbots (Gen AI) and AI-enhanced features built into the SaaS products we already use.1
Most of the products we currently use in business already support some form of in-built AI. I have personally seen demos of AI features in SAP, Salesforce, Microsoft Word, Excel and many more products. The capabilities which AI adds to these systems are already enabled and being built upon. In that respect, AI isn’t just a feature any more - it’s a platform.
This means that you can’t not use AI without disconnecting from the ecosystem your competitors, partners, and customers already inhabit.
Your Partners’ use of AI affects you
Let’s imagine you decide you don’t want to deal with the problems around introducing AI and decide to turn all AI features off.
Well your suppliers and contractors are still use it. You just won’t be aware of it.
In my role as Head of AI at the UK’s largest supplier of gas to business, a procurement lead asked me what our policy should be if someone from another company started an online meeting by stating that AI would be used to record and summarise the meeting. At first that seems like a great idea — it could save a lot of manual labour, and may even be less biased than a human would be when making minutes.
But when you use GenAI for a while, you realise it can make mistakes.
And that means risk.
For example, what if the AI attributed something that was said to the wrong person? What if it misinterpreted a word, then changed the following sentence based on its misinterpretation? Simple errors that could lead to disaster.
Unsurprisingly, we decided there needed to be a policy agreed for all meetings if AI was going to be used to summarise them.
So even if you turn AI features off in all your systems, you will still have to create policies for its use. Your suppliers and contractors are still using AI to generate proposals, code, and documents that flow into your business.
You’ll still be affected. You just be pretending you’re not.
Your Board wants to use AI
I have lost count of the times I have been asked to give someone high up in an organisation an unsupported license for ChatGPT or Microsoft Copilot. They’d heard how “cool” AI was and wanted to give it a go.
Once I had to set up a special training session for the board of directors and C-Suite — a five-figure expense for a half-day session — simply because the chairman was interested in finding out more about AI.
And I have seen Gen AI tools selected based on promises made to the CEO that Microsoft Outlook’s AI can now summarise emails for you.
When someone at that level believes that you can make their lives easier by turning on a new AI tool, it takes a brave person to keep saying no.
Is AI just hype?
Some people are now claiming that we have reached the peak of the “AI bubble”, the top of the AI Hype Cycle, and that soon it will all disappear.
The “hyperscaler” companies - OpenAI, Microsoft, Amazon, NVidia, Oracle, Meta (Facebook), X.com (formerly Twitter) and Alphabet (Google) - have already invested 100’s of billions of US dollars between them in AI. Microsoft alone is projected to spend $80bn on data centers to support AI in 2025.
With all that investment, are any of them simply going to give up?
Imagine that all-but-one of them actually did stop investing in AI. What would the last one standing do? Would they also give up, or would they use the opportunity of a monopoly to invest even more?
After the first internet boom and bust, we were told the internet was just hype and that it would go away. But then came cable modems, wi-fi, 3G, mobile data plans and smartphones, all making the internet more and more accessible. The technology around the internet kept developing, and 20 years later, the internet is something none of us can live without.
It’s the same with AI - it’s not the technology itself that’s stopping it being useful, it’s how poorly it’s integrated into our existing systems at the moment.
It actually doesn’t cost much to run a single AI on a single machine. Anyone can do it for free. So someone will always be willing to pay to develop the technology itself. Even if that’s just a student in a university.
And that means that AI is not going anywhere. It’s only going to get better.
The real questions we need to be asking
So the question whether we should use AI or not is already a long way back down the road. The remaining questions for us as leaders are:
- Where does AI actually add value,
- How do we control it, and
- How do we measure success?
It’s time for you to stop questioning. The decision is made. You don’t need to waste money on more and more experiments any longer. It’s time to focus on fewer, better ones which will start to answer these all-important questions on behalf of your company.
In the next article, we’ll move past the false debate of ‘Should we use AI?’ and focus on what really matters — where it adds value and how to stay in control.
There are also some newly innovative products which have arisen simply because AI has enabled them, but they are still quite rare at the time of writing. ↩︎