Governments, universities, and industry are all talking about the future of work, ethics of algorithms, and biases of data. These talks are mainly among people who set the policy and build the tools. But the voices of those who’ll be most affected are absent. 

It’s policy makers’ responsibility to represent those people and debate the coming challenges. Those that are publicly acknowledged, like job loss, but also those that are hidden, like data bias. They then need to try to find a solution.

Currently, AI’s impact is limited to existing technologies and simple tools that reduce the need for human labour. However, some applications of AI are raising concerns.

• Deep-fake technology in video propaganda can mimic faces and voices with remarkable accuracy.
  
• Self-driving vehicles create legal problems, particularly around responsibility and liability when a computerised driver replaces a human one.
  
• Governments are using AI technologies to advance national agendas or act in protectionist ways. 
  

The crossroad NZ (and the rest of the world) faces is realising short-term gains while not being distracted by these long-term concerns. 

AI is currently used in very narrow use cases. And when done well, most people will not even realise it. 

Take Google Waze, for example. A Waze user can share real-time information about traffic accidents, slippery road conditions, or significant petrol price changes. Its AI system instantly uses that data to adjust local driver recommendations or notifications. To the user, this appears to happen automatically. But behind the scenes, Waze is using machine learning software Anodot to make these suggestions.

This “Narrow” AI is already being used by businesses in a range of applications. But, they each come with trade-offs.

Text analysis

• Xero is using AI-driven text analysis to pull data from PDF invoices for automatic input into their software,
  
• But this could be used for workplace monitoring.
  

Facial recognition

• Google can automatically tag your photos with peoples’ names,
  
• But this can lead to racial profiling.
  

Chatbots

• Online insurance company, Lemonade has used chatbots to vastly improve customer outcomes,
  
• But chatbots can be trained to inherit biased views, as in the case of Microsoft’s Tay experiment.
  

When it comes to emerging technologies, NZ often lags behind the big players. But not for the lack of effort. We have a small domestic market, limited technical expertise, and few avenues for capital. These limitations pose a real challenge. 

So, does this mean AI will only be developed by the big international players? And what about those enormous data sets you need? 

I’d argue that NZ can be a player. Maybe not the leader, but it’s not a race. However, we need to act swiftly, decisively, and with sound leadership. 

The AI Forum’s research project, Towards Our Intelligent Future: New Zealand's AI Roadmap is a good start. It’s a government/industry co-funded research project intended to advance understanding of how AI can be applied to NZ’s business and state sectors. It‘s ultimate goal is to stimulate investment activity in AI.

However, we do have some advantages on our side.

• We have high quality, accessible data sets.
• We’re early adopters.
• Our people have creative mindsets.
• Our small market can help minimise costs.
• And we’re tech savvy. 

That’s why companies such as Google and Facebook use NZ as a test bed for their new features.

The recent AI Forum flagship research paper, Artificial Intelligence: Shaping a Future New Zealand identified that just 36 per cent of NZ organisations surveyed said their company’s board were discussing AI. 

If we’re to remain competitive internationally, NZ business leaders and boards need to start asking practical questions about AI.

• Are we ready to adopt AI? 
  
• What benefits and opportunities can AI bring to our business?
  
• What legal implications do we need to be aware of?
  
• What ethical and social impacts do we need to consider?
  
• What are the competitive threats from AI? Particularly if my competitors adopt it?
  
• Do we have the right data sets? And is the data in a ready state?
  
• Who can help us build AI tools?
  

Choosing to adopt AI is a complex business decision. As such, it should only be made after careful consideration. Seek advice from service providers who have designed and built AI tools in the real world, such as Springload.

Like to understand more about the adoption of AI for your business? Our team at Springload would be more than happy to sit down over a coffee and talk through your needs. Just get in touch.