There is a predictable cycle in technology. Something new arrives, investors declare it revolutionary, consultants produce elaborate diagrams explaining why everything is about to change, and companies start spending money partly out of optimism and partly out of fear of being left behind. Then, after a few years, the mood shifts. Finance departments begin asking difficult questions, executives quietly wonder where the promised transformation is hiding, and a growing number of people start pretending they were sceptical all along.
Artificial intelligence is simply the latest example of the pattern. The excitement is enormous, the spending even larger, yet underneath the noise there is already a growing sense of uncertainty about what exactly businesses are getting in return. That uncertainty is not necessarily a sign that AI has failed. In many ways, it is probably the stage where serious technologies begin.
We have seen this before. The dotcom boom promised to reinvent commerce, which it eventually did, although not before destroying huge amounts of capital and producing business models that should probably never have survived a first meeting (I was there and saw that first hand). Cloud computing followed a similar trajectory. Early promises focused on lower costs and simplicity. What many organisations discovered instead was rising operational spend, greater complexity, and a long-term dependency on a small number of providers with remarkably sophisticated pricing structures.
Even electricity disappointed people initially. Factories replaced steam engines with electric motors but kept the same layouts and processes. Productivity barely improved. It took years before manufacturers realised the real advantage of electricity was not simply replacing one power source with another but redesigning the factory itself around what the technology made possible. Only then did productivity surge.
That is usually the part people forget. New technology rarely creates immediate transformation because organisations tend not to change their behaviour as quickly as they change their tools. Most businesses do not reinvent themselves overnight. They bolt new systems onto old structures and expect dramatic results to follow automatically.
AI is currently being treated in much the same way. Many organisations have inserted it into existing workflows and assumed efficiency gains would naturally appear. Sometimes they do. More often, what emerges is speed without much improvement in quality. Reports arrive faster but say less. Marketing content scales endlessly but begins to sound oddly interchangeable. Decision-making becomes more efficient while at the same time feeling thinner, as though something important has been streamlined out of the process.
The problem is not the technology itself. Much of it is genuinely impressive. The uncomfortable reality is that automation often exposes weak thinking faster than it creates strong thinking. That creates a particular risk for marketing and strategy teams, where the temptation to produce more content, more campaigns, and more analysis can quietly overwhelm the harder task of producing work that is actually distinctive.
This is usually the point where eagerness cools and scrutiny increases. Projects that sounded strategically essential six months earlier suddenly need to justify themselves commercially. Boards start asking about measurable returns rather than theoretical transformation. Quite a few of them struggle once subjected to normal commercial scrutiny.
Oddly enough, this tends to be the most productive phase of the cycle. Once the hype fades, businesses are forced to focus on where the technology genuinely adds value. They redesign processes instead of simply accelerating them. They stop treating automation as a branding exercise and start treating it as an operational tool. They also accept that some things probably should not be automated at all.
That is when the benefits usually become real. Not through dramatic overnight disruption, but through gradual improvements that show up quietly in productivity and competitiveness. Less exciting perhaps, but considerably more valuable.
The strange thing is that we already know this cycle exists, and yet we continue to repeat it. Partly because no organisation wants to be seen missing the next major shift, and partly because the long-term payoff is often real. The internet did transform commerce. Cloud did reshape IT. AI will (almost certainly) change how businesses operate. The mistake is not believing in the potential. It is assuming the benefits arrive neatly packaged and immediately accessible.
They rarely do. Most major technologies spend years looking either overhyped or underwhelming before organisations figure out how to use them properly. That awkward middle period is not evidence that the story is over. More often than not, it is the point where the real work finally starts.
By Peter Zanatta
