Ride the artificial intelligence wave: Four technologies making AI accessible

Artificial intelligence (AI) is no longer pure science fiction, relegated to books, television and dystopian movies - we're well past that point today.

Growth and investment in AI have been astonishing, and its set to continue with Gartner predicting AI technologies will be in almost all new software by 2020. 

While there has definitely been a laser focus on AI in the past few years, consumers and businesses have actually been exposed to AI for a long time, perhaps without even knowing.

As an example, how do you think Outlook knows which emails to put in your spam folder? Or how does your favourite online retailer know which other products you might like?

The technology is undeniably there, and it's becoming more accessible every day to businesses of all sizes.

Consumers are now demanding better brand experiences, as Siri, Alexa and the like become more ubiquitous, with the focus for businesses now on using the AI technologies available to them to dramatically alter the way they engage with their customers.

Amongst these technologies already at the fingertips of companies, four, in particular, will be instrumental in unlocking the true potential of AI.

The early beginnings of The Internet of Things

More machines are now instrumented than ever before, as sensors - categorised as Radio Frequency ID (RFID) devices - have multiplied across all types of places and things in the past decade.

For years, businesses have been using these devices for a number of purposes, one example being package tracking.

RFID chips can set off alarms if customers walk out of a store without having them removed.

That little chip thing? That's an early and crucial element of the Internet of Things (IoT), which – when connected to the internet – can now provide information back to the main system and alert the relevant people.

The information provided by the sensor on its own isn't that helpful.

But what is helpful is lots of that information accumulated over time.

A big use of this is in oil pipelines.

The pipelines will have a sensor every twenty feet or so, and if they start to sense a vibration in one, it might be on a scale of one to ten.

Over time, if spikes are seen at certain times of day or coincide with certain weather activity, adjustments can be made.

One of the key benefits of IoT for businesses is the myriad of data being collected from multiple sources.

Collected data gives businesses the ability to look at data from a distance so they can see the forest through the trees.

First there was data. Then there was big data, and data lakes

The influx of new data-led technologies over the past decade has led to businesses becoming data hoarders, collecting and storing it without really knowing what to do with it.

And this is only set to accelerate with as IoT gains pace.

For all intents and purposes, data storage today is unlimited, but that doesn't mean businesses have to store everything.

One way to combat the data storage issue is with 'data lakes' — a repository where businesses can store any and all data in its original source, and effectively analyse it together.

Data lakes are crucial to any data analytics strategy, and one of the foundations of any well-oiled AI strategy.

Computational power

Addressing big data also requires a large computational infrastructure – essentially power – to ensure data is being processed and analysed correctly.

Fortunately for us, in 2017 computational infrastructure has never been more powerful or readily available.

Part of the reason why industries including manufacturing, healthcare, finance, and business are now able to process massive raw data sets is the availability of cloud computing platforms which can extract the valuable information.

Man meets machine: Machine and meta-learning advances

Today, almost everything produces data to some degree, and now we are better equipped than ever to store, analyse, and process this data.

This gives enterprises the power to create better models and algorithms that improve machine learning, giving businesses new avenues to capitalise on unexplored opportunities.

Businesses can now harness machine learning to boost efficiency, reduce costs and improve insights and results.

They are now equipped to solve a variety of complex problems, from predicting the durability of production machinery to anticipating or even preventing recalls.

Harnessing meta-learning principles –a sub-category of machine learning – can help capture learnings from prior experiments to then be applied for future experiments.

This drastically minimises the need to invest in expensive and hard to find data scientists, allowing organisations which don't necessarily have large IT resources to run leaner and produce even stronger results.

This growing approach is key to enabling organisations across a wide spectrum of markets to implement AI solutions that can produce powerful results.

Artificial intelligence is now a part of our everyday lives, allowing organisations to deliver more convenient services than ever before.

To survive in an AI-led world, businesses need to ensure they are harnessing the power of the technologies they are already using, or which are at their disposal, to develop AI strategies that will ensure they remain relevant and competitive.

Article by Mark Armstrong, Progress international vice president and managing director (APJ - EMEA)