It seems as though artificial intelligence has been on the brink of going mainstream for decades, accompanied by the waxing and waning hype cycles often associated with almost-there technologies. But this time it’s different. This time the hype is supported by real-world applications. Today, enterprise organizations are applying artificial intelligence to real-world applications that are capable of generating significant value.
Many articles dealing with artificial intelligence begin with ironic references to the machines taking over and Skynet, but, the reality of artificial intelligence is more mundane and more exciting for its mundanity. Breakthroughs in neural net research, particularly convolutional neural networks, and other machine learning techniques, coupled with lessons learned from big data analytics, and readily available cloud infrastructure, are capturing the interest of executives across numerous industries. Investment in AI startups has increased by an order of magnitude over the last six years. 81% of IT leaders are investing or planning to invest in AI.
Perhaps the most familiar customer-facing manifestation of AI is AI chatbots leveraging natural language processing. Many banks and credit card providers are using AI voice recognition systems to allow users to carry out simple transaction via interfaces like Amazon‘s Alexa. Santander’s voice banking system will eventually allow customers to fully service their accounts using a voice interface to the bank’s app.
But AI goes much deeper than the ability to handle everyday customer service tasks. One of the more interesting internal applications of machine learning is fraud prevention. Machine learning systems are capable of analyzing vast amounts of data to spot patterns largely invisible to humans. Historically, fraud-spotting systems have relied on human developers to explicitly code the capability to find fraud-correlated patterns in transactions. The difference with machine learning is the systems are capable of figuring out the correlations and patterns without being told what to look for.
Machine learning is also being leaned on heavily in both the healthcare and cybersecurity fields. Many areas of healthcare depend on the ability to process massive amounts of data in search of patterns that provide useful insights for disease diagnosis and preliminary interactions with patients. UK company Babylon Health recently raised $60 million to build an AI chatbot focused on diagnosis.
In the cybersecurity field, researchers are applying the power of machine learning to spot threats that human operators would remain oblivious too. So many subtle factors and conditions may be correlated with an ongoing cyberattack that only machine learning is capable of identifying exploits and vulnerabilities fast enough to make mitigation a reality.
AI-based technologies are improving quickly and companies are moving fast to exploit the commercial opportunities. But for AI to be effective, enterprise organizations have to be ready with the data, the expertise, and the infrastructure to make the most of new technologies. In the coming decade, the competitive landscape across a swath of industries is likely to be shaped by artificial intelligence adoption. Forward looking CIOs and CDOs would be well advised to invest the time to consider how AI might generate value and competitive advantage for their organization.
