Large language models (LLMs), not automation, represent the true inflection point in artificial intelligence (AI).
This is according to Vukosi Sambo, CEO at STM Healthcare, who says language-centric AI has fundamentally changed how intelligence is accessed and applied.
He was one of the keynote speakers at the ITWeb Data Insights Summit last week. He argued that AI only reached mass relevance once it became conversational. For decades, AI remained confined to specialist domains, limited by technical interfaces and narrow use cases. That changed when LLMs aligned AI with the most natural human capability – language.
“Civilisation itself began with language,” Sambo said, explaining that every major leap in human progress, from speech to writing to digital records, has been driven by improved ways of capturing and sharing knowledge. LLMs, he noted, reconnect intelligence to that same foundation.
Sambo identified 2022 as the moment AI crossed into mainstream use, driven by the release of conversational systems such as ChatGPT by OpenAI. For the first time, intelligence could be accessed through dialogue rather than code, dashboards or specialist tools, he noted.
This shift unlocked unprecedented adoption, he said. "People don't need to learn how to use AI. They already know how to talk."
While organisations are generating vast volumes of data daily, Sambo warned that data alone has no value. The real challenge lies in synthesising information fast enough to support decisions, something humans can no longer do unaided.
LLMs, he argued, should be viewed as decision-intelligence engines, capable of processing complexity at scale, providing context and improving judgment rather than simply generating content.
LLMs are rapidly democratising access to intelligence, lowering barriers that previously favoured technical specialists and well-resourced organisations, Sambo asserted. However, this also reshapes competition.
“As access becomes equal, advantage shifts to context, governance and data,” he said, warning that organisations that fail to embed LLMs responsibly into their decision processes risk falling behind.
Even if the AI market experiences corrections, Sambo believes the language-driven shift will endure. Once intelligence becomes conversational, it cannot be reversed.
“This is not a bubble that disappears. It’s a structural change in how humans interact with information, and it will shape how we work, compete and make decisions going forward,” Sambo said.
Share