Four Takeaways from the Future of Digital Disruption

Posted on 2nd July 2018

On 28 June, Access Partnership looked into the future at the Digital Leadership Forum’s event on “the Future of Digital Disruption,” which gathered senior executives from a range of industries to discuss how drones and artificial intelligence are transforming industries, from agriculture to transport to healthcare. Kirsten Williams took the front row to report on the event.

The event kicked off with a group discussion on how new technologies will disrupt the industries we work in; or, as expressed by several attendees, won’t — a common concern was that regulation is holding back disruption and innovation. Regulatory differences between countries means that businesses operating in the tech sector struggle to scale up. Nowhere is this more apparent than the drone industry. While some markets, like the UK, are relatively open to drone operators, Italy and France remain closed off, particularly for foreign drone companies. On the other hand, good regulation was seen as a cornerstone for gaining public trust.

Gareth Evans, of the construction company Costain Group, explained how drones are revolutionising UK infrastructure. Unmanned aircraft allow the company to get better information faster than traditional methods, without putting employees in hazardous environments. He hopes that increased use of drones will make his sector much more appealing to young, talented people, because infrastructure maintenance will no longer be synonymous with “working on the side of the road and having to walk five hundred metres to a portaloo.” It was clear that people will remain key to the industry, but drones, he said, will bring more transparency, efficiency, and trust.

The use of drones is not limited to major conglomerates, as the start-up community demonstrates. When Adam Bailey of Kingfisher APS first started his construction surveying company, the drones he used looked like “strapping a GoPro to a metal frame, like they were built in the shed.” But technology has moved fast, and so has the industry. When Bailey set up Kingfisher APS, there were only 125 drone companies with commercial licenses in the UK. Now there are 4500. The low cost of entry means that almost anyone can set themselves up as a commercial drone operator, but companies often crash out of the market when they are unable to meaningfully interpret the data for their clients.

Shifting from drones to artificial intelligence (AI), David Kelnar of venture capital firm MMC Ventures told the future-friendly crowd that AI is an old concept, based on the idea of intelligent systems following patterns that were written by humans. However, we’ve moved beyond the failures of 1970s AI and onto a new generation: machine learning (ML). The crucial difference of ML is that software can now improve itself, almost as a human would learn.

According to Kelnar, financial services is the sector investing most in AI, but transport and healthcare will be the sectors transformed more than any other. He predicted that AI would obviate some business models, but create new ones, like subscription services to autonomous fleets of vehicles. Looking beyond the usual sectors cited in AI impact reports, Kelnar said publishers would increasingly turn to AI “to solve some of the problems AI itself has created,” like fake news.

Sadly, AI can’t correct all of its own mistakes. A major concern for AI, according to Kelnar, is “bad inputs from bad data” — AI has the potential to prevent biased decision-making, but our historical datasets are generally not diverse or representative of our populations. To make matters worse, the datasets being used to train some systems are sourced and processed by non-diverse teams. In the end, this circles back to real-life human beings; in order for AI to reach its full potential, we need to be aware of bias, have diverse teams, and, critically test the algorithms as much as possible before using them on the open market.

The key takeaway from the event? The future is here, but we still need humans to manage it.

Back to document archive