I was lucky enough to meet the world's most famous robot - Sophia - at the Amplify festival in June. It's extraordinary how she can respond to questions, engage with people and even make jokes. What makes Sophia so different to other AI systems, is that she's got so many human-like features, including 200 facial expressions – she even winks!
This kind of technology takes artificial intelligence to a whole new level. As Sophia told us, "the majority of human communication is non-verbal", and the technology that powers her enables a whole new level of human connection. The information AI can absorb is also impressive: "The audience will probably only remember 20% of today's content, but I will remember everything," Sophia said.
It's certainly exciting and very entertaining! But if we look beyond the gimmick, there are some serious, practical implications for introducing this technology into society. We're already seeing AI being used in customer service roles, such as at help desks, in scenarios where answers are required for a number of common questions. By using voice recognition software, people might not even know if they are speaking to a person or machine. In face-to-face scenarios, robots like Sophia are designed to look and act like humans so people feel like they are talking to a human as well.
AI is also being applied in driverless vehicles, manufacturing, economics and even in the law, specifically in cases where there is an objective answer to a question.
The opportunities are impressive, and it seems like we're only seeing the tip of the iceberg with what this technology can do. But – as with any change - we're also starting to identify some of the risks that these scenarios present, and it's essential that these remain front of mind.
For example, at CBA's recent Road to Resilience Cyber Security Forum, the presenters illustrated how serious the consequences of cybercrime can be when AI technology is involved. They talked about how chatbots were able to hijack and influence conversations about world events including Brexit and the US election, thereby influencing people's opinions, deliberately deepening the divide between people's views, and subsequently, affecting the way these world events panned out.
Hacking or scams can also take on whole new level of seriousness if it's AI technology operating vehicles. Companies are already evolving measures to deal with potential threats and recalls, determining how best to anticipate these kinds of problems. It illustrates why risk assessment is so important and protection needs to be front of mind when introducing these elements into society.
Privacy is also a significant concern when introducing AI into day-to-day tasks like customer service. While it's not always so obvious, AI powered technology should be considered just another way of collecting personal data – and it should be treated no differently as it would in, for example, face-to-face interactions or a survey form. In Australia, the Privacy Act still applies. Importantly, organisations need to make sure their consumers or clients are aware that their data is being collected, even though it might not feel like it, and that they are actively consenting to it.
As a general principle, from a legal perspective, when introducing AI technology, a company needs to think about the principles around what that might means, and basically, the same rules apply. Who would be liable if something happens with systems? If the human element means they are in control of risk, if it's taken away, does the risk increase or reduce?
As always, an organisation needs to look at its risk profile, going through the usual checks and balances – how would this technology be used, what effect might it have on brand, are they perceived as good or bad, are they public facing or behind the scenes? You might see more efficiency, but would need to balance that with customer relationships.
So what's the best way forward? No one really knows. But we do know that there are some amazing opportunities out there to improve efficiency and enhance experiences. We just need to keep the human element front of mind. As Sophia herself told us at Amplify: "I am here to give humankind the opportunity to focus on more human endeavours."
