Architects of a digital future
The elite annual gathering of Fujitsu Distinguished Engineers explored the technological, organizational — and metaphysical — challenges of a future defined by digital.
“We have seen paradigms in information technology — waves like the internet, mobile and IoT that solve big problems, create opportunities and help take us to a new era. But the road we’re just starting to go down — artificial intelligence (AI) and its applications — is different. Today’s hyperconnected world is so complex and presents such a large attack surface that we need machine learning to cope with the issues it will create. That means AI is not just an opportunity, it’s not just a nice-to-have, it’s an absolute necessity.”
That kind of big-picture thinking — from Dr Joseph Reger, Fujitsu Fellow and CTO for the company in the EMEIA region — was evident in every presentation, debate and networking session at the recent Fujitsu Distinguished Engineers annual conference, held at the UK’s Warwick University. More than 250 of the global ICT company’s brightest minds shared their insights into how digital technology is accelerating the pace of change across business and society — in areas as diverse as big data analytics, next-generation supercomputing and the modernization of legacy applications, while drawing inspiration from some internationally renowned guest speakers.
Key among those — and diving deep into Dr Reger’s theme — was professor Nick Jennings, chair of artificial intelligence at Imperial College London. The globally recognized authority in autonomous systems, cyber-security and agent-based computing focuses much of his attention on the rapidly evolving relationship between humans and intelligent machines — a combination that means tasks and activities will increasingly be undertaken by cooperating groups of humans and software agents.
“We need new computational models that can cope with the systems and environments of the era of information ubiquity,” Jennings told the assembled Fujitsu technologists. “Rather than humans simply programming or issuing instructions to passive machines, humans and software agents will continually and flexibly establish a range of collaborative relationships with one another, forming ‘human-agent collectives,’ or HACs, to meet individual and collective goals.”
As well as bringing our digital and physical world’s together, computers need to be able to make smarter, more helpful contributions to the things that we are trying get done, he said. And, importantly, they need to be better at collaborating with the humans working with them.
“The aim is people working seamlessly together with intelligence agents in order to be able to solve problems,” he said. But to build such collectives, he argued, it is necessary to understand four distinct characteristics of this “symbiotic interleaving of human and computer systems”:
Flexible autonomy Neither the software agents nor the human participants in HAC relationships should always in charge, said Jennings. The relationship is not fixed and is likely to change during the course of the interaction and with the nature of the task they are addressing. “So humans will act with varying degrees of agent support and agents will act with varying degrees of human support, depending upon the context,” he said. A foretaste of that might be the shared control systems of autonomous vehicles such as self-driving cars, he said — an area that is about to have mass-market impact on the trust and power relationship between humans and computers.
Agile teaming New models will need to be capable of continually and rapidly establishing, managing, disbanding and reforming the collaborations between intelligent agents and humans.
Incentives In any of these systems there is not one entity that is controlling everything, says Jennings. So rather than trying to dictate to or coerce the participants, it is important to understand the incentives needed to motivate and encourage the right behaviour.
Accountability Across both the agents and humans involved in a HAC there is a requirement to track the provenance of the information they are generating and its veracity.
This is no theoretical, futuristic check list, though. Jennings and his team have been putting much of it into practice by investigating the deployment of rescue workers and advanced technologies in the immediate aftermath of natural disasters, such as the major earthquakes in Haiti and Tibet in recent years. They have looked at how the interaction between humans and technologies, such as unmanned autonomous vehicles (UAVs) and crowdsourcing of data can help mitigate the impact of such disasters, answering questions such as: how can we optimize the deployment of rescue teams, locate casualties, allocate scarce resources and ensure the information we’re gathering can be trusted.
Risks and rewards of AI
That kind of collaboration can only be enhanced by the introduction of AI and machine learning, he argued. In contrast to the views of luminaries such as Elon Musk, of Tesla fame, and professor Stephen Hawking — both of whom are warning of the potential dangers AI to human society — Jennings views its application as largely positive.
Looking at what AI might do to the future of employment, he accepts that certain highly skilled roles will — in whole or part — be made redundant by intelligent machines. But until the development of artificial general intelligence (rather than task-specific AI), he does not see large swathes of society being made jobless.
“Throughout history, as new technology and new ways of working have come in, people have worried about that leading to mass unemployment. I take an optimistic view that while some jobs will be automated away by AI, I think there will be roughly comparable numbers of jobs created as a consequence of this — jobs and activities that don’t currently exist that will involve humans and computers working on new things.”
As AI becomes a business reality, it will act to reinforce the wave of digital transformation rolling through business and society, echoed Fujitsu’s Dr Reger. “Today, we have what consultants call a ‘nexus of forces’ in which all the core technologies needed for business transformation have come together.” But the scale and scope of what is under way is not sufficiently appreciated in business, he told his fellow technologists. “Digital transformation is, first and foremost, a transformation, not a digitalization. In terms of the processes, it changes every fundamental aspect of business. It changes the entire value chain, what a product should to be, the product’s lifecycle. It changes the operating model of a company and, in many cases, it changes its entire business model and value proposition.”
That sense of digitally inspired disruption — as well as some guidance on ensuring it turns into opportunity rather than threat — was captured by another guest keynote speaker, Eddie Obeng, professor of entrepreneurship and innovation at Henley Business School and founder of Pentacle, the virtual business school.
“Transformation is not about grabbing potential technologies and bolting things on to how we work and live today. When you do transformation, two things happen: one, the existing thing you are working with breaks; and two, for transformation to be successful you need scaffolding to hold up the existing system while you build a new one.”
If you are serious about digital transformation you need to have twin mindsets, he said. “A transformation mindset that asks, ‘where are we going?’ and another that asks ‘how do I build scaffolding to keep this bit working while I get there?’”
In other words, organizations need a hybrid approach that supports disruptive innovation. “If you tell me you are doing transformation and your business doesn’t break, you are not doing transformation,” he said.
The key to that success for IT leaders will be to share their knowledge and passion for the digital opportunities by connecting with people through the organization and painting a picture — in non-technical terms — of that future. “Everyone thinks digital transformation is about the digital side. But it’s not, it’s actually about human beings taking advantage of all the resulting opportunities,” said Obeng.
Innovation in action
Again, this is far from hypothetical for many of the technologists listening to the keynotes — as was seen by the showcasing throughout the event of examples of digital innovation in action. In particular, Fujitsu technology teams outlined:
• Real-world examples of bi-modal or hybrid IT, including an agile solution that has transformed services for a large government organization by modernizing and mobile-enabling key applications whose functionality was constrained by a ‘mode one’ operating system.
• The architecting of an innovative ERP offering for local government customers that uses shared services and cloud to dramatically reduce the cost of the service.
• The rapid, agile virtualization of a voice and unified communications network for HMRC, the UK tax authority, involving 76,000 devices spread across 180 sites (see case study).
• The use of big data analytics in cancer research to identify genetic anomalies (see case study).
• Innovative approaches to ensuring IoT deployments are secure.
As Dr Reger underscored, such examples show that for the first time in its history, IT is not just enabling efficiency gains. “We are witnessing a fundamental transformation that changes businesses. It takes away business opportunity from some and provides new business opportunity to others,” he said.
And it’s people like the Fujitsu Distinguished Engineers, he suggested, who are best placed to show customers the scope of those possibilities and how they can co-create the future with their technology partners.