2022 may be the first year in which industries begin the slow move to Industry 5.0. But as 2021 comes to a close and employers, tech visionaries and the wider world look to innovation in 2022, what can we all expect from that most opaque, yet revolutionary of advancements: AI?
As we catapult from the depths of COVID-19 into a future where we are rapidly incorporating the best and most complex elements of our fourth industrial revolution, AI sits on top of a throne of innovation.
While the contemporary use of AI is limited by the tech of today, the opportunities and potential of AI to transform the way we live, travel, shop and communicate is unrivalled.
But AI is still an evolving, highly unknown, cultural and computing niche – what AI could do for humanity in regards to the transference of masses amounts of labour, service provision and services from human-operated and created, to robotics and machine learning would upend centuries of societal requirements and the value and meaning of life, work and money.
There are, of course, plenty of public, and famous, uses of AI that have crept into the mainstream as a universal “good”. These AI systems are driving discoveries on far off celestial bodies (such as the AI tools on the Perseverance Rover on Mars), to powering the tech now powering delivery drones and autonomous driving (Tesla leading the way, but other companies such as Apple and Google are not far behind). AI is slowly becoming ubiquitous. But as with anything tech related, there are costing, infrastructural and sustainability limits which have to be overcome to build fully integrated, societally accepted and easy to incorporate AI systems that benefit us all.
So what can we expect from the business of AI in 2022? What are the next small victories in the industry that will drive innovation for future businesses? Will any of these advances actually make any difference to our way of life straight away?
- Transformer Machine Learning
To the layperson and general consumer, this may not have a huge amount of impact on our way of living, but from the perspective of machine learning it’s enormous: the ability of deep learning systems to not have to learn in order of data input, but on relevance of input using something called Attention Mechanisms.
This then creates more effective and more efficient AI systems, specifically improvements to niche programming systems such as Natural Language Processing. Transformer machine learning is sometimes referred to as a new architecture for machine learning. In other words, it’s a huge advancement.
- AI and Climate Change
AI is a vital tool in our efforts to stem the use of carbon-intensive products and build a more sustainable society. From using satellite imagery to spot methane leaks to improving efficiencies in factories, manufacturing plants and carbon-heavy industries, AI is making huge waves in changing processes and improving efficiencies.
But it’s also helping in a more holistic, data-led way by providing more advanced simulations and models that can help climatologists predict more extreme weather patterns, or discover new methods of carbon capture.
- AI Patents
In most countries only Humans can patent new creations. However, this year South Africa granted a patent to a creative AI system – a world first. Although this is outlier AI use, it’s the first example of just how important AI systems can be in our creation and discovery of new medicine, products, systems and solutions.
- AI Inside
This refers to “AI that’s embedded at the core of everything from architecture to operations”. This means AI systems embedded within devices (such as phones, fixed servers, digital devices etc) to bridge the gap between data, device and user and create more immediate and relevant AI decision making. This means AI is closer to us, the user, and therefore can provide more digital tools, help, services or alerts in the moment.