How Honest is Your Vendor on its AI Capabilities?
‘AI washing’
Building and marketing AI will be high on the agenda for all smart enterprises, but some may be jumping the gun and marketing themselves as having AI services when they don’t. Analysis Jim Hare from research and advisory company Gartner, backed this view in a research piece ‘How Enterprise Software Providers Should (and Should Not) Exploit the AI Disruption’. He said, “Similar to greenwashing, in which companies exaggerate the environmental-friendliness of their operational practices for business benefit, many technology vendors are now ‘AI washing’ by applying the AI label too indiscriminately.” Some businesses may be mistaking its rules-based engines for AI systems; just because the engines have 100s of rules, doesn’t mean they are AI. Words such as AI, machine learning and rules-based systems are now commonly thrown around and it can be confusing. However, there are clear differences. To differentiate these models, a clear definition of what AI is, is beneficial.
Defining terms
Rules-based engines are a traditional way to build expert systems to automate decisions. They use a series of programmed ‘if-then’ statements that guide a computer to reach a conclusion or recommendation. It is based on two main components; firstly, a set of facts about a situation; secondly, a set of rules for how to deal with the facts. AI and Machine Learning are terms often confused with one another. AI is defined as the simulation of human intelligence processes by machines, especially computer systems. It is the broader concept of machines being able to carry out tasks in a way that we would consider ‘smart’. Whereas, machine learning is an application of AI that provides systems the ability to automatically learn and improve from experience without being explicitly programmed. Machine learning focuses on the development of computer programs that can access data and use it to learn for themselves. Stanford University defines machine learning as ‘the science of getting computers to act without being explicitly programmed’. You need AI researchers to build the smart machines, but you need machine-learning specialists to make them truly intelligent.
Which is best?
Each system comes with pros and cons. Rules-based engines know what to expect; they have expertise given to them based on past experience. But rules are only as good as the programmed research. One way of looking at it is to think of rules-based engines as being in the past and AI being all about the future. For example: AI with machine learning used in cyber security will say, ‘I don’t know what this is, but there’s been something similar before, so I’ll flag it.’ Or, ‘I’ve never seen this before, it’s an anomaly; I’ll flag it.’ AI recognises anomalies merely because they are new. Whereas, rules-based engines will use if-then decisions based on past data. Ultimately, a rules-based engine will never improve by itself; it needs someone to constantly update the rules. Whereas AI improves its accuracy the more it is used; the more you use it the better it becomes.
Making sure you get what you pay for
When speaking with vendors about their AI capabilities, make sure to ask how their ‘AI’ systems deal with the unexpected. You can’t write rules for the unexpected – there is no history. Therefore, the answer they provide you with will uncover whether their product is truly AI or rules-based. SCC is a long-term advocate of the use of AI. Our Public Safety Video Analytics Service utilises artificial intelligence and deep learning algorithms to identify potential elements of interest in large amounts of footage. Discover more about SCC’s services here