AI-DAY 2018: AI In New Zealand is Alive and Kicking

Quick Take

The AI-DAY event in Auckland on 28th March 2018, presented by the AI forum and NewZealand.AI, illustrated a creative and motivated AI solutions industry flourishing in New Zealand. The job for New Zealand organisations is to determine their AI strategy, to mitigate competitor disruption and to understand and manage AI implications to society and the economy.

Event Highlights

The Minister for Broadcasting, Communications and Digital Media, opened the AI-DAY event to a crowd of around 500 attendees saying that New Zealand needs to accelerate its thinking on AI to grow the economy and protect citizens. "AI has a huge part to play," she said about kickstarting New Zealand's digital economy. Curran noted that the government is already involved in AI related fields. The Ministry of Social Development is developing a human rights and ethics framework that will have implications for AI. The Ministry of Business, Innovation and Employment (MBIE) is assessing the implications of AI on jobs. The Privacy Commissioner is investigating issues that arise from the use of algorithms.

The event showcased some exciting New Zealand AI solutions focusing on autonomous vehicles, natural language processing, image recognition and geospatial data:

•  Ohmio Automation Autonomous Shuttle Buses. Mahmood Hikmet from Ohmio explained 'localisation'; a robot's understanding of where it belongs in the world. At Ohmio, their self-driving vehicles use "sensor fusion" where several sensors, such as GPS, odometry and LiDAR pinpoint the vehicle's location. Ohmio is trialing its shuttles on predetermined routes at Christchurch airport this year.
• Re:Scam. Powered by IBM Watson natural language understanding, Re:scam sends intelligent replies to email scammers. This wastes their time and increases their cost to profit says Netsafe's Sean Lyons. Re:scam has several inbuilt profiles that it can select to reply with different emotional responses. Replies will demonstrate frustration, anger, excitement, surprise or interest, for example, at the appropriate time. The solution uses human typing traits, such as typos, to throw scammers off the scent that they could be talking to a bot.
• The Cacophony Project. Predators in our forests and wildlands are reducing the native wildlife in New Zealand. The Cacophony Project aims to reduce predators and increase native wildlife using AI. Cameras capture infrared images of moving objects and AI identifies if the moving object is a predator, such as a rat, possum or stoat, or another object such as a bird, human or a pet. Once identified, the predator can be eliminated. Speaker Grant Ryan said that the solution lures predators by recognising their calls and responding. To measure their success, the cacophony project uses AI to measure the changes in birdsong to measure the repopulation of native birds.
• Orbica. The fascinating intersection of geospatial data, imagery and AI. The tricky part is how to represent, visualise and extract the intelligence. Founder Kurt Janssen says that, for example, councils could use residential area data to determine the number of unconsented buildings or unfenced swimming pools. An Orbica solution monitors construction progress at Thyssenkrupp's construction sites. Drones capture visual images which are 3D modelled with geospatial data. AI identifies construction features such as scaffolding, building framing and cladding, to obtain a picture of construction work. It compared the picture to the previous week's images to determine that week's progress. You can play with Orbica GeoAI data at this link

Perspective

At the moment, in New Zealand, end-user organisations are exploring AI. AI encompasses a wide range of technologies and processes and there is no single definition that can truly encompass it. But, this does give AI a broad appeal, with opportunities across every vertical. Because AI is more about solving problems than about technology, it appeals to the business as well as to the IT department. Speakers at the event gave a number of messages around understanding AI for the enterprise, including:

• AI is about Man plus Machine, not Man versus Machine. Westpac's CIO Dawie Olivier said AI is about augmenting human intelligence, such as calculating all the parameters to give good financial advice. Where AI removes the need for humans to perform tasks, it frees humans up for higher value tasks.
• Where does AI fit in? Steven Guggenheimer, CVP Microsoft, explained that building AI solutions is about building a layer of intelligence above the system of record (where data comes from) and below the system of engagement (UI's for webpages, social media etc). This placement of the system of intelligence enables learning loops between data, users and AI.
• Data is an input to thinking. But it's not the be all and end all, said panelist and Figure.nz founder Lillian Grace. Data is the underlying input to AI algorithms but often data gets used for purposes that it was not collected for.
• Context is Crucial. Being able to act with intelligence requires contextual knowledge and understanding, not just generic knowledge stated Westpac's CIO, Dawie Olivier. This will be the difference between automating something and informing something.

The challenge for adoption of AI is that organisations aren't prioritising talking about Artificial Intelligence in the boardroom or executive office, said AI-DAY panelist and founder of The Mind Lab, Frances Valintine. The lack of strategy leaves New Zealand organisations at risk of disruption by global competitors. 

We only have to look at Amazon as a classic example of using AI to disrupt competitors. Amazon uses AI across its business, for example, item recommendations, robots that pick shopping baskets, packing and dispatch automation, predictive shipping and more. Amazon's recent launch in Australia stirred concerns over the eventual fate of local brick and mortar stores (or even local online stores), if they can't compete with Amazon, in part fueled by artificial intelligence. If AI driven disruption can happen to your local bookstore or appliance store, then it can happen in your business too.

New Zealand organisations need to put AI at the top of the tech priority list in the boardroom and executive office. Companies must start understanding the implications of AI across their sector. For societal and economic implications, this may require strategic level collaboration in the private sector with involvement from the public sector. Individually organisations must consider the commercial and competitive implications of AI and develop an enterprise-wide strategy. Expect the pace of change and adoption of AI to move fast.