Most AI pilots can look impressive in a demo, but proving their value in the real world is a different challenge. This is the gap NTT Data Imagine event addressed.
It took place on 27 October at The Forum in Bryanston, under the banner Architecting AI-powered enterprises to deliver impact.
The evening opened with keynotes and a panel. Guests then moved through hands-on hubs, and the event travelled to Cape Town the next day.
NTT DATA Imagine 2026
Marc Jessiman, MD of Technology Solutions at NTT DATA SA, opened the evening with a story about his nine-year-old son. He hoped companies would build a kill switch, and he wondered if humans could get a chip in their brains.
To succeed in this era, Jessiman said, we need to reinvent ourselves and stay curious. Creativity and courage matter too.
The gap between stimulus and response
Futurist John Sanei argued that AI demands creativity, imagination and curiosity from us. Businesses cannot simply add AI to existing operations and expect different results. Their business models and organisational structures need to change, too.
Instead of CEO-led companies, he pictured ‘flatter’ organisations with small teams directing thousands of AI agents to solve problems continuously. But adapting to this shift also means letting go of familiar identities and ways of working.
Sanei described three phases of change: sadness, strangeness and adventure.
- The first involves leaving behind what we know.
- The second is an uncomfortable period when old assumptions no longer hold, but the way forward remains unclear.
- Adventure comes when we begin to find clarity.
He says most of us are still somewhere between sad and strange.
For Sanei, the answer is not to compete with AI on intelligence. Just as we don’t compete with a tractor, our role is to direct it. That requires human qualities such as judgement, experience, empathy and imagination.
He pointed the famous Viktor Frankl quote about the gap between stimulus and response as the space where those qualities emerge.
“Between stimulus and response there is a space. In that space is our power to choose our response. In our response lies our growth and freedom.” – Viktor Frankl
When we constantly react to the world, we fall back on familiar patterns and repeat yesterday’s ideas. Making better decisions requires us to pause, question our assumptions and imagine different possibilities.
ALSO READ: Stolen passwords are SA’s biggest ransomware risk in 2026
Fix the process first
Bill Wilson, Global AI Leader at NTT DATA said businesses should first ask whether a process is designed well before adding AI agents to it.
He gave the example of a UK government AI solution handling full protection orders in the criminal justice system. Before its implementation, these orders were often delivered to victims’ home addresses instead of the local police station.
After the solution went live, 50% more orders reached the correct destination, helping address a potentially dangerous flaw in the process.
His advice? Double down on one or two key value chains and show the value of AI there.
Wilson also cited a survey in which 95% of respondents called sovereign and private AI important, yet very few have acted. He blamed complexity, since your own AI means stitching together many layers and partners.
A PS5 in a data centre
The PlayStation 5 display also grabbed my attention, because duh.
NTT DATA has one sitting in its Johannesburg data centre, with a local console on site for comparison. The remote PS5 ran Dirt 5 through Sony Remote Play, over NTT DATA’s software-defined core network, with the picture reaching the screen via LTE and a breakout port.
The demo came with a racing wheel setup, so naturally, I gave it a go. Driving games are very much my weakness, though, and I know my limits. Give me an RPG or a shooter any day.
The team then demonstrated how quickly it could change the connection’s bandwidth through its client portal. Starting at around 50 Mbps, they deliberately reduced it to 4 Mbps, degrading the image on screen and causing packet loss.
When they moved the slider back up, the network applied the new configuration within about two seconds. The graph displaying packet loss took longer to catch up, with a delay of around five minutes.
The team can use API calls to automate network changes. NTT DATA is also looking at MCP communication, which could let external AI agents communicate with agents inside the network, with people setting the guardrails and budgets.
ALSO READ: Shadow AI: New tool to map unsanctioned AI in real time
The Riyadh AI Factory
NTT DATA also announced the global expansion of its AI Factories. These hands-on environments help clients test AI workloads against real-world requirements, including infrastructure, performance, security, governance and cost, before moving into production.
The Riyadh AI Factory serves as NTT DATA’s strategic hub for the Middle East and Africa. It launched in September, followed by facilities in Paris and Singapore. The three sites join the inaugural AI Factory in Plano, Texas, and the Munich facility, which launched on 7 October.
Dilip Kumar, President of Technology Solutions at NTT DATA, Inc., said: “AI at scale is ultimately about turning intelligence into measurable business impact.”
