These tools, and many others, show that the capabilities of AI have increased exponentially in the last few years. And this wave of innovation is poised to hit the business world with a new generation of AI-powered apps set to redefine what’s possible.
The most transformative are those that enable “real-time” AI. All AI is powered by information (data), and real-time data is the most valuable data. After all, what’s most helpful in making decisions is knowing what’s happening now, or what happened a week ago?
Some of the most useful and well-known AI-powered applications have found success with real-time AI. Think Netflix, which uses up-to-date information to recommend movies and shows to watch next. Or Uber, which uses real-time AI to match passengers and drivers as quickly as possible. Even Google and Amazon use real-time AI to put ads in front of the user and recommend products to buy based on his latest shopping trends.
I talked about the latest version of my podcast Ed AnafChief Product Officer of data stacks – Helps companies operationalize data and make it work in real-time AI applications. We talked about how companies that can move quickly into the real-time AI space will not only be able to take the lead from their competitors, but also be able to meet and exceed customer expectations.
What is real-time AI?
Today, when we talk about AI in a business context, we usually mean machine learning (ML). ML includes algorithms that become increasingly efficient at doing their jobs as they are trained on increasingly large datasets. Real-time AI refers to ML algorithms that can process and learn from streaming data captured as it happens.
Some examples are:
Banks and credit card companies monitor financial transactions as they occur in real time to identify potentially fraudulent ones.
E-commerce businesses such as Amazon analyze customer behavior in real time to offer the products and services they are most likely to want to purchase at that moment.
Online advertising platforms such as Google and Facebook analyze what users are searching for in order to provide more relevant ads that are more likely to result in sales.
Manufacturers of self-driving cars monitor activity around their vehicles to help them navigate safely to their destinations while avoiding hazards.
Cybersecurity professionals analyze network and email traffic in real time to try to predict where hacking attacks and phishing communications are coming from.
Anuff tells me: arrival? ‘How long does it take for Uber to arrive?’
“Real-time means doing this continuously, based on all the data events that are happening. [right now] …real time means not after the fact. ”
Where does real-time data come from?
Business activities and interactions can be decomposed into events. Each of these events, if captured in a way that can be analyzed, can become part of a real-time streaming dataset.
This includes everything from credit card transactions, clickstreams of user interactions browsing websites, to data collected from RFID scanners and computer vision-equipped cameras in stores and on the street.
Once businesses are able to collect these data points, they can be fed into ML algorithms to identify patterns and trends. These may be too subtle to be detected by human analysis, but they provide valuable insights that can be used for prediction.
“And when you tie it to those interaction points, you can focus your attention on what actually impacts your business. You can…have a human come to that conclusion, so the business is optimized in all sorts of amazing ways.”
Other sources of real-time data ripe for use by AI-powered businesses include:
Social Media Activity – Businesses can use it to track mentions of their brand and direct marketing activity in real time.
Sensor data from Internet of Things (IoT) devices – such as tracking human movement and enabling predictive maintenance of tools and equipment.
Transaction and POS data – Identify fraud and identify upsell and cross-sell opportunities.
Environmental data such as weather, humidity and pollution levels, and economic data such as stock prices and exchange rates – understand how they impact markets and customer behavior.
Why is this important for businesses now?
Actually this is nothing new. The examples discussed so far show that technology leaders like Netflix, Uber, and Amazon have been using them for decades. What is relatively new is that an ecosystem of machine learning and analytics providers, and the platforms that enable them, are now within reach of nearly every business.
Anuff said: [but] These tools and databases were not really available. ”
This is the result of an ongoing phenomenon called the “democratization” of AI. Technology that was once the domain of Silicon Valley’s tech giants and scarce academia is now packaged, commoditized, and available to all. Anuff tells me it mirrors what happened before in the democratization of other technologies, from the dawn of mechanization to the advent of computers and the Internet.
“This is a pattern that we see time and time again when there is a new technical capability that is only understood by experts at first… but over time, mere mortals can do it and use it for a variety of purposes.” It will be democratized until the application.
Anuff is particularly excited about how the convergence of different techniques and forms of AI analytics will lead to revolutionary new applications that will revolutionize what is possible in the near future.
A good example of this is the combination of image generation technology and natural language processing. This has enabled applications such as Dall-E to take written statements and transform them into images.
“It’s like an iceberg, most things are under the surface…In the world of machine learning, a lot of things have been going on under the surface for a long time, but they’re suddenly unleashing…People are very Be creative and put things together in new ways.”