Signpost: The revolution will be generic

Five enterprise technology conferences. Five audiences running into the tens of thousands. But by the third keynote, the speakers have blurred into one collective voice reciting the same laundry list of AI agents, orchestration platforms and trusted data frameworks.

IBM Think, Microsoft Build, Adobe Summit, ServiceNow Knowledge, VeeamON – and that is before the data and creative platforms join the chorus. Close your eyes during any keynote and you are adrift in the same sea of hybrid infrastructure, sovereign environments and trusted data frameworks. Open them again and you are none the wiser about which message came from which stage.

A decade ago, this would have been unthinkable. IBM events had more gravity than bankers reading nuclear engineering manuals. Microsoft was the productivity giant eager to be cool. ServiceNow specialised in making corporate workflows marginally less soul-destroying.

Today they are all building the same machine. Or, at least, that’s how it sounds.

The labels still vary, naturally. IBM speaks of “AI operating models”. Microsoft prefers “the agentic web”. Snowflake talks of “the AI data cloud”. Veeam talks of “data resilience for the agentic era”. But that is branding rather than differentiation. Strip away the corporate packaging, and the active ingredients are identical: AI agents, orchestration, governance, automation, live data, compliance, visibility. Swap the logos and conference delegates would not raise an eyebrow.

In other words, enterprise AI has entered the generic-drug stage.

New technologies follow a similar path to patented pharmaceuticals in their early years. Wrapped in exclusivity and premium pricing, vendors insist their formula is unique. Then the market matures, ingredients standardise, and terminology converges. Buyers stop asking which vendor has the most compelling vision and instead ask whether the product integrates with procurement systems and complies with European regulations.

The most telling aspect of the current conference circuit is what companies no longer say. Few speak of invention or of a fundamentally different vision. The language has, instead, migrated: from possibility to plumbing.

What happened?

AI turned into infrastructure.

Electricity was once magical. So was mobile telephony. So was cloud computing. For goodness sake, so was WhatsApp, at least in messaging-hungry countries like South Africa, India and Brazil.

When the novelty fades, buyers refocus on reliability and cost. Nobody chooses a power utility because its electrons are visionary. Enterprise AI is heading down the same road, and that is bad news for marketing departments that have spent the past two years treating every AI update as if it were the invention of fire.

There are already signs of fatigue inside large organisations. The appetite for the new has not disappeared, but the tolerance for promises that do not reach the end of the food chain is wearing thin. Most non-executive employees are still trying to get expense systems to approve an Uber claim. The disconnect is impossible to paper over with keynote pizzazz.

This partly explains why the current generation of conferences has shifted in tone. IBM presses on governance and compliance, offering organisations a sense of control over AI decisions. Microsoft pushes integration, offering enterprises already embedded in its ecosystem a path of least resistance, even as the CEO calls “Code Red” on Copilot’s failure. Snowflake positions itself as the neutral ground where data from all these systems can finally speak to each other. Adobe, meanwhile, reassures creative and marketing teams that AI is an assistant rather than a replacement.

Underneath all of these lies the same inconvenient truth: the products are becoming harder to distinguish, and technology is no longer the main differentiator. When that happens, distribution networks and installed customer bases take over.

Exactly like generic medicine. Enterprise customers are finding themselves making the same calculation with AI that they make at the pharmacy counter.

And the 2024 conferences are calling to ask for their demos back. Buyers now want compliance, stability, compatibility, governance, integration and, critically, somebody to blame if things go wrong.

Those generic enterprise drugs favour incumbents. Who owns the office environment? Who carries decades of enterprise credibility? The winners in the generic AI era may not be the companies with the smartest models. They may simply be the companies the tech doctors prefer to prescribe. For now.

In the generic-medicine market, nobody asks which factory made the tablets. They check the dosage, the price and the safety record. Enterprise AI is now shopping at the same counter.

The revolution has not been cancelled. It has simply gone generic.

* Arthur Goldstuck is CEO of World Wide Worx, editor-in-chief of Gadget.co.za, and author of “The Hitchhiker’s Guide to AI – The African Edge”.