At this point we’re all familiar with the predictions. The dataverse – the amount of data created each year – will, it is confidently expected, reach 175 zettabytes by 2025. It was 33 zettabytes in 2018. It has also been predicted that, in the same time frame, there will be some seven internet-enabled devices for every man, woman and child on the planet, all of them generating data. Partly, the growth will be driven by the desire for connectivity and accessibility in less-developed countries, with a particular spotlight on the African nations.

But even as these predictions were being made, the tech landscape was changing. There’s been a decade of talk of the Internet of Things – and how it will transform everyone’s lives. Likewise AI has been a topic of wide discussion – a buzzword – for as many as five years. As we enter the third decade of the 21st Century, however, we have now reached a tipping point. Things are happening and now that they’ve started, they’ll happen ever more quickly.

Instead of AI – which implies a machine mind that passes the Turing Test – consider Machine Learning, or ML. The global ML market was valued at around $1.58 billion in 2017 and is expected to reach approximately $20.83 billion in 2024. Machine Learning enables software applications to be more precise in predicting results without being definitively programmed.

There are some who would have you believe that by 2050 all intellectual tasks performed by humans could be accomplished by artificial intelligence (ML) technology. Common applications for the technology are already found in the fields of financial services, healthcare, government, marketing and sales, transportation, oil and gas, manufacturing, bioinformatics, computational anatomy. In fact, there are very few aspects of day-to-day-life, business and personal, that aren’t (or won’t be) touched by ML tech.

Just in case there’s anyone left who might be tempted to consider this as simply a passing fad, the market uptake of this technology is creating as much fear as it is opportunity. Around the world, governments are implementing process to enable united and collaborative decisions on how AI (ML) might be used now and in the future.

In terms of societal impact, children born from 2010-2025 are part of what’s called Generation Alpha, and are considered to be the most tech-infused demographic to date. In 2017, a report stated that the new generation would be much more aware of their AI interactions.

It said: “They will converse with digital assistants, learn new skills from robots and be driven around in cars that are controlled by AI. Generation AI will become more independent as they grow up, thanks to assistance from AI, which will actually force them to become interdependent on the technology.”

In 2020, when the oldest of Generation Alpha are 10 years old, there’s already a mushrooming of the amount of devices, applications and innovations that make use of machine learning, and the IoT that enables machine-to-machine communication.

This year’s Consumer Electronics Show, January 2020 in Las Vegas, provided a flavour of what the new generation can expect:

  • Smartypans has a frying pan that checks the weight and temperature of ingredients before guiding you through the cooking process via an app.
  • Inirv will swap your cooker’s knobs for its smart dials.
  • PantryOn aims to automate food shopping lists via smart shelves that monitor when a family’s favourite groceries run low.
  • BrightLock unlocks front doors by detecting a pattern of light pulses fired from a smartphone’s flash.
  • Amaryllo’s Athena is a security camera that recognises people’s voices and faces, to distinguish friends and family from strangers.
  • Elephant Robotics showed MarsCat, a robo-cat that can play with toys, recognise its owner’s voice and even interact with real cats. It is programmable and can be used to teach students to code AI applications.
  • Tombot promoted its robotic labrador puppy designed to provide comfort to residents in old people’s homes, and others who would benefit from a pet, but cannot deal with a real animal.
  • Yukai‘s Bocco robots offer a way for children to send and receive voice messages to their parents, and babble back in their own language if addressed themselves.
  • Human Capable’s Norm glasses will be able to make calls, show directions and recipes, and both shoot and play videos – so a less geeky-looking Google Glass for the 2020s.
  • Waverly Labs has a second take on language-translating earbuds.

The implication of all this is that the predictions are understating the case – there will be far more data created than was expected in 2018. The dataverse potentially will be much, much bigger than we believed. In this scenario, demand for data centre infrastructure, and the power and connectivity that it requires, will increase beyond what is already expected and planned for.

Without the capacity to store, manage and process the data, the revolutions promised by the IoT and the applications of AI (or, more accurately, ML) won’t happen. It’s not a question of whether limits can be put on data centre expansion, nor is it a time for NIMBYism.

It’s about sustainability in data centre construction and operation. It’s about renewable energy sources, increased efficiencies in energy usage and innovation in areas such as cooling and on-site power generation.