The AI invisibility cloak
The benefits of AI are already building. Why are they invisible? Where should you look in your business to find them and what comes next?
There is a persistent theme emerging in commentary about AI in business. The central thesis is that AI projects are failing and the benefits are not being seen in key economic numbers such as productivity.
Around this are clustered doubters who see signs of the end of the hype cycle. On the other side optimists are convinced that, in the words of Zhou Enlai, “it’s too early to say.”
Which of these groups is right? Or are we asking the wrong questions and talking about the wrong things?
My belief is that the change is already underway and the value of AI is being realised every day. It is just hidden behind an invisibility cloak.
Le Chatelier’s principle
Systems bite back. Coming from fundamental chemistry but with surprisingly wide application. Complex systems have invisible automatic stabilisers. Improvement (or deterioration) in one factor will be balanced by a reaction that keeps the overall outcome stable.
Business systems work in a surprisingly similar way. Individual improvements in productivity are often absorbed into buffers and balances within a system. Less scientific analysts might call this the giant hairball of company processes. Or refer to its close cousin the slow moving train wreck project.
So just because profits don’t move doesn’t mean there is no value. Most likely the value is ignored or invisible.
Business change projects are carefully planned and structured to achieve well defined outcomes, often labelled with the unconsciously Orwellian acronym SMART.
If the benefits don’t fit within this neat box of expected outcomes, they can safely be ignored or attributed to other causes. It’s budget season. All our customers are focused on that new regulation. People in Guildford bought more ice cream this month. You know the sort of thing.
That is if they are visible at all. Those SMART objectives are embodied in carefully selected metrics. The benefits may not be reported simply because everyone is measuring the wrong things.
Note that this is happening even at the macro level. There is a lot of focus on the absence of changes in productivity. You will see less commentary from experts about the impact of AI on GDP.
This is because AI is making an impact on GDP, at least in China and the US. In exactly the way economic theory predicts. Right now, GDP is increasing because of massive investments in AI infrastructure.
You don’t have to look far to find extraordinary charts showing the levels of this investment.
Where to look
Yet somehow all this is invisible? So where could you look for the business benefits of AI today:
Corridor whispers. Who is talking about it? How are people saving time or producing more research/ greater detail/ longer analysis?
In the shadows, on the blurred edges and between the cracks. For all sorts of reasons, people may not be open about their use of AI today. Security policies, organisational culture and just a sneaky feeling that using ChatGPT is somehow cheating. Formally and informally, encourage openness and have a learning mindset.
At the bottlenecks and jams and overflows. Speeding up some tasks can cause overload at different steps of the process. Look out for new complaints about delays or backlogs in unexpected places. Maybe some of the benefits are lost in those holdups.
In your inbox, social media feeds, networks and all around you. Don’t just look at how your team is using AI. Observe what is happening elsewhere. Chances are someone in your company is doing the same things others have found successful.
What should business leaders be doing with AI today?
Do nothing
Despite what you may hear, this is always a legitimate option. Personally, I think AI is too exciting to leave alone. Not everyone is as much of a tech optimist as I am.
Do what everyone else is doing
There is an emerging consensus approach around cautious pilots, proceed only when ROI is proven, careful controls and guardrails plus “human in the loop”. You will not get fired for doing this but you won’t see much benefit either.
Explore and learn
This would be my preferred approach. Look at what your people are doing and see what works for them. Observe the world around you. Almost everyone you meet is using AI or trying to use it. What can you learn from them?
Go all in
Maybe you are in a business where AI is already making a difference. Customer service, some areas of software development and parts of marketing are good examples. If your industry is already going there, you may need to bet everything now. I don’t think this applies widely but it will be exciting and fun if you are at the leading edge.
The invisibility cloak will be around for longer than you might think. The benefits of AI will keep accruing underneath its protection and they will still be hard to see. A few pointers about what might come next:
Flocking
“The value of connectivity to farmers is broader than any single task or action. Connectivity unlocks vast opportunities that were previously limited or unavailable,” Aaron Wetzel Vice President at John Deere.
Your people are not operating in isolation. They are talking to colleagues, sharing with friends, discussing AI in the pub or on the golf course. This is how human progress works. We are individually quite smart but our collective mind is exponentially more powerful.
(Incidentally, this is why comparing AI to the intelligence of individuals is pointless and irrelevant.)
Over my career, software has gone through three cycles: product, service, network. AI will follow the same trajectory.
Collaboration and competition will emerge. New projects, communities and markets will form. How will your business be positioned in the new landscape?
The investment cycle
Logically, all that infrastructure investment will start to drive prices down. Lower cost for more AI capability is likely to unleash a new wave of innovation. People will talk about startups again. Who will your new competitors be and what will they offer to your customers?
That cycle may also lead to a bust in the stock market and beyond. Many commentators think this is already imminent. They may be right. That will not change the medium term potential for your business. Giving up on AI when things go wrong would be a mistake.
Beyond Le Chatelier
I mentioned Le Chatelier’s principle earlier. That works within the boundaries of a system. It falls apart when the boundaries and limits that define the system are breached. In science and nature, that requires evolution.
In business, evolution is much faster. As new forms of collaboration and competition take shape, the internal structure of organisations will stop making sense. Beyond that, boundaries between companies will become blurred. Before you know it whole industries are disrupted.
The AI change is already underway. Your people are finding ways to use AI that generate value. Engaging with AI to help with things they care deeply about. You are probably doing the same.
Summary
Explore where the real benefits are emerging. Observe what is happening around you, inside and outside your business.
Learn what works and encourage flocking.
Don’t do what everyone else is doing. Look underneath the invisibility cloak and find the opportunities AI offers for your business.
Thanks for reading