Predictions 2018: Artificial Intelligence (AI) & The Internet of Things (IoT)
By Dr Joseph Reger, Fujitsu Fellow and Chief Technical Officer in EMEIA at Fujitsu
- Ubiquitous virtual assistants – Over the course of the year we’ll see an increase in the number of virtual assistants and chatbots that are currently in use by banks and insurance companies. Consumers will notice a significant ramp over the next year; however by 2020 they will no longer notice that their interface is not human.
- Government health programs will turn to AI – We’ll see governments start to invest in AI to improve health and reduce medical spending. The focus will be on leveraging the technology to accelerate diagnosis and for preventative medicine.
- At least two disruptive AI business models will emerge in the next 12 months – We are currently witnessing digitalization disrupt many industries. We can expect to see the start of a new wave of AI-based disruption – with retail a likely candidate for significant change.
- Eighty percent of all larger organizations will investigate AI – Sixty percent of these will actually be undertaking proofs of concept. When they reach the implementation phase, 100% will encounter a lack of appropriately skilled employees.
- Over the course of 2018, the net effect of AI will be positive for the workforce – a whole new market for AI-based jobs will emerge, and the technology will not yet significantly replace workers. However, as AI systems increasingly automate many traditional jobs – this will create a growing challenge by 2020.
- By the end of 2018, artificial intelligence will be used to optimize the next generation of production AIs – to date, techniques where one AI is used to improve the performance of another have only existed in research settings. In 2018 we will see such techniques, e.g. Generative Adversarial Networks (GANs), deployed to enhance production AI systems.
- All aspects of manufacturing will use AI to some extent by the end of 2018 – all manufacturing companies will use AI for at least some part of the value chain either in logistics, manufacturing or maintenance.
- Assembly line workers will increasingly have robotic colleagues – to date, robots have been restricted to working inside cages on manufacturing processes, however, the emerging generation of smart autonomous robots can see, touch, and collaborate safely with humans while taking on the heavy lifting of assembly work in addition to boring, routine tasks.
- Increasing conversations around the ethics of AI – we’re going to see a marked rise in debate regarding the ethics of AI. Concerns will be openly debated at government level in a number of countries.
- AI systems will start to become virtually invisible – as we rely on AI for more and more tasks, they will rapidly become as familiar as all the other technologies we interact with every day. This growing reliance on AI will lead to ‘artificial’ intelligence being accepted as ‘natural’.
- Masses of IoT data to sift through: In 2018, substantially more IoT data will be collected than we can possibly analyze and understand. We have yet to fully understand the importance of deleting transient data that may not ever have any value.
- Machines first: In an increasingly hyper-connected world, 2018 will be the year when IoT devices start to focus more on becoming attractive towards other smart products, rather than humans. Any IoT device that makes it hard to exchange data with another machine has a limited outlook.
- Rapid IoT consolidation: Today, there are approximately 900 IoT platforms. Not all will survive the year as global standards emerge and drive out proprietary and stand-alone platforms.
- The year of IoT security: Even simple sensors will be able to provide secure communications and any insecure IoT project will have a short lifespan. The need for secure communications between machines on the industrial shop floor – both internally within an organization and between producers – will drive the acceptance of global security protocols. The need for secure and undeniable transactions will mean that blockchain will become more than a digital currency.
- B2B is driving IoT, not B2C: In 2018, we will continue to see investment and progress in business-to-business IoT deployments by far outstrip business-to-consumer projects. IoT will become embedded into B2B products at a pace faster than the consumer market can keep up.