Martin Dippenaar, CEO of Global Kinetic, unpacks why frontier intelligence will demand significantly more tokens than current mainstream applications.
As it currently stands, most organisations are still at the experimental stage when it comes to adopting AI. While many are looking to leverage agentic solutions within their environments, the majority of organisations are still unclear about what their ultimate objective is when it comes to this pervasive technology.
This experimentation has given rise to a token economy, which is the metric by which AI usage and cost is measured within businesses.
As organisations become more savvy in terms of how they deploy AI, understanding which types of models can be applied to specific business processes, will result in better spending. While cost is expected to decrease, there is still a looming bill on the horizon, according to Martin Dippenaar, CEO of Global Kinetic.
Citing a recent Gartner report, which has predicted that the cost of using LLMs will be 100 times more efficient by 2030, Dippenaar highlighted that organisations wanting to make use of the latest models, will still need to pay a premium.
“Frontier intelligence will demand significantly more tokens than current mainstream applications. Agentic models, for example, require between 5-30 times more tokens per task than a standard GenAI chatbot, and can perform many more tasks than a human using GenAI,” the Gartner report outlined.
“While lower token unit costs will enable more advanced GenAI capabilities, these advancements will drive disproportionately higher token demand. As token consumption rises faster than token costs fall, overall inference costs are expected to increase,” its analysts added.
As Dippenaar pointed out, while AI tokens will become more affordable, those organisations wanting an edge will need to be operating at the frontier level, and this comes at a price.
“On the surface tokens may be cheaper, but a ‘good’ AI output today usually means bigger context windows, more data stuffed into every request, multiple model calls chained together, and often a higher‑tier model in the loop. So yes, the price per token is down, but the number of tokens per serious task has exploded, and expectations of quality have gone up at the same time. The net effect, for real production workloads, is that the true cost per useful answer is often flat, or even higher than it was just eighteen months ago,” he explained.
As such, those organisations wanting the latest and most accurate information, need to prepare themselves for the cost it will carry.
Looking at the South African context, Dippenaar offered some insight into the industries that would be most affected by the token economy, as well as noting the sectors that are better insulated from such costs.
Regarding the latter, he looked at SA’s robust fintech and banking sectors, most of which are environments where governance is critical and embedded deeply into the operational model. “Highly regulated sectors, like banking, lean towards committees, pilot teams and policy documents rather than large‑scale deployments. The result is that the true cost of AI is still mostly theoretical, with models still being tested and not yet embedded into core systems at any significant scale,” he shared.
While these types of organisations are okay in the short-term, in the medium term, Dippenaar noted that when businesses in this sector finally switch on AI at customer scale, the surge in token usage, along with the associated costs, could be significant.
Identifying other industries where AI tokens are expected to have a massive impact include call centres, where these businesses handle millions of interactions in a week, all of which are potential candidates for agentic AI. “Each of these interactions can silently fan out into multiple prompts, long contexts and expensive model tiers, so that a small increase in cost per conversation can turn into a very large monthly bill,” highlighted Dippenaar.
Given the current state of flux that AI has wrought, as well as how rapidly the space is evolving with each day, very soon, organisations will have to make some important decisions, particularly when it applies to the cost of hiring a human to do a task versus the cost of AI tokens. Here, Dippenaar refers to a crossover point, which could occur within the next two years locally.
“As token prices fall but model size and usage intensity rise, providers will be forced to keep the effective cost of an AI agent below that of a human agent. If they don’t, companies in call‑centre‑heavy industries such as airlines, telcos, and banks will find the economics of full automation far less attractive than the marketing pitch of AI,” he emphasised.
“Companies are right to make the most of AI. We are using it to great effect in our own work, and even guiding clients on how best to optimise it. But it’s vital to remain clear-eyed about the hidden costs, before they wipe out any potential gains,” he concluded.
[Image – Photo by Eduardo Ramos on Unsplash]
Get the tech news you want to read. Take our reader survey and tell us how we can help you better.
Take the survey now
Follow the story
