Programme

There is an overwhelming amount of information (and misinformation) about COVID-19. How can we use AI to better understand this disease? In this session, we take an open dataset of research papers on COVID-19 and apply several machine learning techniques (name entity recognition of medical terms, finding semantically similar words, contextual summarization, and knowledge graphs) which can help first responders and medical professionals better find and make sense of the research they need. We will dive into the techniques used and share the code repository, so developers will walk away with the understanding of how to build a similar solution using Cognitive Search.

The 4th Industrial revolution arrived at speed when we weren’t even talking about it during the start of the Pandemic era. Of course we were more focussed on looking after the safety of our families, friends, communities but the world was suddenly digital. Sectors like Health and Education went digital overnight, eCommerce rates soared, WFA was invented and massive industries like tourism and aviation were wrecked. What happens next, are we even ready for a digital world that is seamless, frictionless and virtually borderless - are we ready to seize the opportunity, compete and grow or are we going to get stuck in the pre pandemic era and face a more rapid decline as we get closer to 2030?

Can AI be used to better protect wildlife and bring accessibility to those who need it the most? Research teams at Google are attempting to do so. Automatically analyzing audio data at scale using Google AI can help biologists monitor wildlife more efficiently. For example, to automatically detect the presence of humpback whales underwater. Similar AI techniques are used to enable automatic speech recognition, which can help people with disabilities communicate more easily and gain independence. Join Julie Cattiau from the Google AI team to learn more.

Enthusiasts see a world of opportunity for the use of AI across new domains, but somehow this hasn’t happened yet. This talk will discuss some of the issues that inhibit people from using AI to the full, and ways that these can be overcome over the next few years. The presentation will also propose ways to communicate the benefits and limitations of AI to people who have real problems to solve and limited time and budgets.

Three years ago, Spark’s nascent AI work was taking off, with the team hard at work developing Spark’s data ecosystem and machine learning models. However, the business wasn’t taking full advantage of the insights and opportunities that were available as a result. Today, Spark realises significant commercial value from its AI work and has senior leadership buy-in to rollout of AI across the enterprise. Kallol Dutta, Data and Automation Lead from Spark will describe how the AI team went from standalone team to valued business partners.

The Government Chief Data Steward (GCDS) supports the use of data as a resource across government, setting the strategic direction for government's data management, and leading the state sector's response to emerging data issues. At the heart of this work is transparency and trust. Since the creation of the role in the 2014, the GCDS has partnered with agencies to: • increase their capability to manage and use data • identify and remove roadblocks to accessing data • implement data standards • use new methodologies The question now is whether the same tools, or something altogether different is required to support the AI Sector. What role can the GCDS play in building a thriving and innovative AI industry in Aotearoa New Zealand?

Australia has well recognised the promise of AI to change our lives for better. In 2019, Australia hosted its first National AI summit – Techtonic – which brought together over a hundred experts from across the country to discuss the future of AI. The launch Australia’s AI Ethics Framework laid the foundations for a responsible, ethical and human-centric AI future, and Australia’s AI Technology Roadmap identified three high potential areas for Australia to create jobs, drive economic growth and improve the quality of life of current and future generations. In 2020, the Australian Government Department of Industry, Science, Energy and Resources worked closely with industry, researchers and the community to test ideas and inform the development of an Australian AI Action Plan. Tim’s session will walk through the experiences and lessons learnt along Australia’s journey. It will highlight the role of industry, academia and community engagement to progress Australia’s AI agenda, and how opportunities to work with businesses and international partners have helped to shape a trusted, responsible AI future.

You will be able to book your roundtable the week before the Summit. The options include - Investment in the AI Economy / Human-centred and trusted AI / Our Place in the World / Uniquely New Zealand / Preparing the Workforce

You will be able to book your roundtable the week before the Summit. The options include - European Approach to AI – what lessons for New Zealand? / How to enable AI-driven use cases in organisations / How can NZ Government and industries create a data ecosystem that supports the future of AI? / Considerations for NZ tailored AI systems / Developing a dynamic AI research platform and partnerships in Aotearoa / The role of AI in enhancing sustainability

In New Zealand, 40,000 – 50,000 hectares of forest is harvested each year with most areas replanted while some areas are converted to another land use. Ministry for the Environment conduct deforestation assessments every two years to monitor these changes and meet New Zealand’s international reporting obligations. In 2020 the Ministry turned to an AI based approach through Lynker Analytics to conduct an assessment of deforestation that had occurred during 2017 and 2018. Using an aerial survey, Lynker Analytics captured photography of almost 7,500 distinct areas of potential forest loss in all regions of the country. The imagery was then classified into land cover classes such as cutover, plantation seedlings, pasture, and mature native forest within each target using a Machine Learning (ML) approach. From this a multi-criteria analysis was used to assign each area of possible forest loss a dominant land cover and replant status. The model was used as the primary monitoring system to detect deforestation and re-planting and flag those targets to the Ministry. The automated monitoring system proved reliable in detecting deforestation, re-planting and other land cover changes exceeding one hectare. It also enabled more rapid assessment of replant status used by the Ministry for reporting.