In the 18 months since the launch of ChatGPT introduced the concept of generative artificial intelligence (GenAI) to the world, most business have played with the technology, many have trialled real-world applications, but so far few have gone all-in on using GenAI across their organisations.
However, that is set to change as more of the IT budget for innovation goes into GenAI – or so says Abhijit Dubey, the new global CEO of NTT Data.
Appointed in June following the merger between NTT Data and NTT Ltd – he was previously CEO of the latter – Dubey now runs a $30bn tech services and consulting business, a subsidiary of Japan’s NTT Group, a $100bn conglomerate that’s the world’s fifth largest telecoms company.
He says there are three main trends driving technology investments in business – and that’s where the rise of GenAI is having a growing impact.
“When I talk to clients, there are three patterns,” he tells Computer Weekly. “[One is] ‘optimise my legacy’, because 70% of costs are still running legacy. Another is, ‘How do I get 30% more productive?’ What we’re seeing is a massive consolidation of vendors to do that. If you take any large organisation, most of them are trying to sort out two to five strategic vendors to be able to optimise the legacy,” he says.
“Then, [the last pattern is] in terms of modernising core systems – do you have the skills, do you have the servers, the tools and the assets to be able to do this at pace and deliver the quality that is required? In terms of building the new, that’s where there’s been a slowdown.
“A lot of companies that have invested in ‘digital’ are asking lots of questions about return on investment, so that [spending] couldn’t go to discretionary products. And I think most of that has now gone into GenAI.”
Gen AI use cases
Dubey sees three stages of progression for organisations considering how GenAI could be used.
“Number one is [companies] experimenting in a bunch of different places, saying, ‘We’ll see what happens’. Number two is, ‘We figured out two or three areas, the ‘no brainers’, where there’s a lot of business value creation, and you don’t need to do a lot of discovery work, so let’s just do it’,” he says.
“The most advanced, which I would say is less than 10% of the organisations we interact with, have figured out there’s a basket of 20 to 30 use cases that will drive the most value, and have a roadmap to lay out how to deliver these use cases over a period of time.
“That involves a new way of thinking and a completely new operating model, architecture and set of providers to work with, because you need a lot more niche skills than you ever [think you] did.”
But there’s a lot more to a successful IT strategy than all the fun new technologies such as GenAI. Computer Weekly met with Dubey on the day the faulty CrowdStrike software update caused chaos around the world. The CEO highlighted some initial lessons from the impact of the bug.
“The technology enablement of the economy has produced incredible gains, and there is no doubt that will continue,” he says. “I can’t speak for the entire industry, but it’s our responsibility to make sure that whatever we design, implement, deploy and manage is the most resilient, and in the event of failure, that we have a recovery mechanism that’s world class so you can minimise the impact of any of these events. This notion of responsibility and resiliency is something that we need to, as an industry, take seriously.”
Unintended consequences
The CrowdStrike situation was an example of the unintended consequences of technology – for all the benefits of digital transformation, it’s almost impossible to plan for every possible negative outcome. But Dubey says the industry can become better at mitigating against the unexpected.
“As technology deployers, what we can control is to make sure that if we have implemented and deployed something, we have done the best we can to make sure it’s trusted and secure,” he says.
“Can you cover every single possible scenario? Probably not. But have you given it your best shot in every single instance and not compromised because of cost and the pressure on economics of running a business? That is something we can control – to make sure that it’s not about putting my company profits ahead of what’s important.”
“What we can control is to make sure that if we have implemented and deployed something, we have done the best we can to make sure it’s trusted and secure”
Abhijit Dubey, NTT Data
NTT is hugely well-known in Japan – for a UK audience, Dubey describes it as the Japanese equivalent of BT. His involvement with the firm goes back beyond when he joined – for 10 years as a McKinsey consultant, he advised NTT on its global expansion, before being brought in to lead that for its NTT Ltd subsidiary three years ago.
Following the merger and the newly expanded organisation he leads, Dubey is focused on correcting the perception that it’s an enormous global player that few have heard of.
“We are pretty invisible globally, outside of Japan. I’d like us to be a lot more visible,” he says, citing the benefits of the entire portfolio now available to the ‘new’ NTT Data.
“There’s no other company that has what we have – everything from physical infrastructure to network, datacentre, public, private and hybrid clouds, security, applications, data, GenAI, business processes, consulting, systems integration services. We can do all of that at scale globally. We’re probably the only company that has that in terms of a full-stack transformation portfolio.”
Full-stack transformation
That’s all very well, but few IT leaders these days are looking to one vendor to provide everything they need. What does that portfolio mean for the reality of today’s IT infrastructure? Dubey describes an example of the sort of project he feels the company excels at.
“When I say ‘full stack’, no client buys a full stack. What it does mean is that for specific domains, you can bring a full stack of capabilities,” he says.
“[For example], any transformation on the edge – a factory, warehouse, transportation hub, mines, and so on. You cannot drive a digital transformation at the edge without having connectivity. If you don’t have reliable, ultra-low latency, high-bandwidth connectivity, nothing can happen.
“In a large factory, Wi-Fi doesn’t work in reality – if you want to drive complete automated operations in a factory, it’s not possible. We bring our private 5G capabilities and our mobile network operator capabilities globally, and we can solve the connectivity problem.
“We also have edge compute that we can deploy, and we can deploy a real-time analytics platform – a smart data platform. And you can deploy that at scale and speed and enable digital transformation at the edge. You could do AI at the edge as well. This is a full-stack solution to enable that.”
For many large IT providers, it’s been a tough couple of years, as demonstrated by the scale of job losses across the tech industry. But Dubey is confident that things will pick up once companies become comfortable with GenAI.
“There are green shoots, I would say. We’re seeing the large transformational deals, especially on the outsourcing side, are back, but the discretionary projects are still slow and they have been for upwards of 18 months,” he says. “Until people really get comfortable with GenAI and the GenAI use cases get deployed at scale, I don’t think that is going to materially shift.”