AI and the cloud: Separating fact from fiction

By Jill Macmurchy, New Relic

From virtual assistants to chatbots and deep learning, AI is taking up increasing space in the cloud. If you believe the hype, the advent of artificial intelligence is destined to make humanity lose control, and will see many human roles replaced by advanced robotics. So, what will the future look like? While no-one has a crystal ball, we can look at what’s currently happening as an indication, and separate hype from fact.

Hype: The cloud knows you better than yourself 

Deep learning is enabling AI to recognise and interpret images as well as text. Computer-vision technology is advancing rapidly; recognising people as well as a variety of categories of objects. Machines can view and process images at lightning speed compared to humans, picking up details – including emotions – that we don’t often have the time to notice.

Fact: Facial recognition doesn’t equal human understanding. Even with medical imagery, AI won’t be diagnosing patients, but instead augmenting doctors’ ability to find the data they need quickly and easily. The cloud is used heavily by the medical imaging sector, but AI is far from its sole purpose.

Hype: AI is set to replace humans in the workforce

There’s rising panic over the “end of work”, with calls for Universal Basic Income to be implemented as an army of robots takes over every job. One theory even speculates that someone could invent AI that improves and replicates itself: creating a new form of life that outperforms humans. Bill Gates and others are calling for a tax on robots in response to future mass unemployment.

Fact: AI does enable companies to save time and human labour when it comes to certain tasks, such as processing big data. The truth is, it will replace many jobs, and is already doing so. But this doesn’t necessarily mean the human workforce will become redundant. For example, the Commonwealth Bank’s online chatbot Ceba is able to recognise 60,000 different information requests, and assists customers with over 200 tasks. Rather than replacing the need for human team members, these employees are able to work on other customer support tasks at the bank.

Hype: AI and blockchain will cost jobs in the banking sector

There are reports that 30 per cent of banking jobs will be threatened by AI and blockchain – up to six million jobs over the next decade. Financial institutions are making huge investments in emerging technologies, increasingly looking to automate and take humans away from the front line of customer service queries.

Fact: The banking sector is rapidly adapting and embracing disruptive technologies in order to stay relevant amid sweeping industry reforms. To not do so puts institutions at a competitive disadvantage. Simply put – some jobs will go, and have already gone, but others will be created.

Hype: AI will become super intelligent and eliminate the human species

The Matrix, Terminator and other sci-fi films put the notion of “machines taking over” into public consciousness long before anyone had even heard of the cloud, AI or IoT. The idea is that machines will eventually “learn” that humans are inefficient and destructive, and terminate the human race for the greater good.

Fact: The reality is that there are issues of ethics and bias in intelligence. Self-driving cars have to be programmed with instructions on what to do in the event of a potential collision. The car may have to “decide” between hitting a single person or a group of people. This is already happening, with Mercedes saying that its future autonomous cars will prioritise passenger safety over pedestrian safety. A hospital life support machine may have to choose between overheating and causing a fire; potentially endangering hundreds of people, or switching off and causing the death of the patient it supports.

Decisions made by AI can sometimes be hard for humans to understand. Computers can assess many variables independently of each other in a way that humans simply cannot. This is already affecting existing technologies – just take a look at the moral bias behind your search results.

Cloud computing is used to host a huge number of applications, with AI tools and devices just one of these applications. With the world changing rapidly, and AI playing a key role in this evolution, we shouldn’t expect a kind of Domesday scenario just yet. It’s up to industry leaders and governments to ensure there are appropriate checks and balances carefully built into systems, and that there is transparency around those checks.

Jill Macmurchy is New Relic Technical Manager for Asia Pacific. She has nearly 20 years’ experience working in leading technology companies, including IBM and SAS Australia.

This article was first published by Computerworld