It's only three years ago since those involved in developing artificial intelligence (AI) saw the enormity of opportunity it offered for human health, according to Haytham Assem, a chief scientist in Huawei Ireland Research Centre.
“In 2018 the AI community started realising the potential impact of AI in healthcare after ProFound AI, a cancer-detection software which successfully assists radiologists in finding early digital breast tomosynthesis (DBT) in patients. This marked what is to come in the healthcare world of AI.”
The car industry had already cottoned on. In 2015 an Audi SQ5 equipped by Delphi Technologies with autonomous driving kit took a road trip across the US in nine days.
“The trip marked a huge advancement in the future of driverless cars and made more obvious the potential impact in the automotive sector for the next decade,” says Assem.
He believes the automotive sector now offers a clear example of how AI will complement and change the way we live.
The industry is currently at what is called “level 2” automation, that is, partial driving automation.
“Soon we are expected to realise level 3 automation which is conditional driving automation in which people inside the vehicle can start engaging in other activities without the need to supervise the technology. There is no doubt that we will realise significantly more adoption of level 2 automation in 2022 and we will start seeing level 3 automation coming into action from 2023 onwards.”
As self-driving car technology becomes more mature, the logistics sector will be a big beneficiary, especially as it has already been disrupted by Internet of Things (IoT) sensor-based technologies.
“By having AI coupled with IoT we are foreseeing that AI will not only be used for determining changes in demand behaviours and optimising inventory levels and replenishment plans to feed the continuous loop of product life-cycle management, but with the maturity of self-driving technology we expect to see autonomous delivery robots ready on our streets in the next five years,” says Assem.
Things are moving on in healthcare too. “AI is already changing the patient and clinician experience in many ways. However we have not realised its full impact by any means. The challenges to date have been more related to the framework for regulatory authorities to approve AI-based applications, and technically closing the gap in the area of “explainable AI” which will help clinicians and patients to understand the reason behind the AI systems’ suggestions,” he says.
Both areas are maturing relatively fast and hence he predicts an increase in AI systems and in crucial areas such as robotic-assisted therapy within the next five years too.
Significant breakthroughs in natural language processing (NLP) technology will result in a rapid growth in virtual assistants across various sectors, including social services, education, healthcare and customer support, he predicts.
These include NLP breakthroughs such as “transformers”, new deep-learning architecture that aims to mimic the human capability in understanding text by being able to remember the context mentioned earlier in a conversation, leading to more reliable applications such as chatbots, language translators and search engines.
Huawei’s Ireland Research Centre is active in this area. “We have built natural language understanding engines that were basically able to understand the intention from what the user is asking in more than 50 languages, with a better performance than human performance,” he says.
“We are hopeful to share our technology soon to the open-source community so they can benefit from it and foster the AI applications market eco-system in Ireland and Europe.”
A host of other trends will impact on our daily lives too, suggests Alessia Paccagnini, academic director of the Master in Quantitative Finance at the Michael Smurfit Business School, UCD.
“The most important trends we are likely to see in this space include facial recognition, which is already used in smartphones or other devices to improve data protection. It will be an important element in our future life for accessing digital healthcare and financial services, and when there is the necessity to have a further level of privacy and data protection in cybersecurity,” she predicts.
She also predicts the intersection between Internet of Things and AI will grow, into AIoT (artificial internet of things) – think Alexa or Siri.
For her DARQ materials are also on the rise. The acronym, coined by Accenture, stands for distributed ledger technology, of the kind that supports cryptocurrencies, AI, extended Reality and Quantum Computing, which together will improve the ways in which we store and process big data, and the quality of information that results.
“For example, the extended reality will improve collaboration from different locations and avoid frequent travel. I could pretend to teach at UCD Smurfit to my students while I will be in New York to talk at a conference or vice-versa,” she says.
She also believes that advances in NLP will improve communication between humans and robots, and that AI-powered devices will help improve services
“For example, in finance by providing new digital services that will substitute the physical banks. In healthcare a glove is in development to help people with Parkinson’s regain a steady hand,” says Paccagnini.
Robotic Process Automation (RPA) will aim to automate business processes while hyper-automation or hyper-personalisation will improve marketing and increase loyalty, allowing for a personalised experience for consumers, she says.
Look out too for GANs, or generative adversarial networks, says Assem.
“These are also perceived to be a promising future for deep learning with their amazing ability to create visuals and images that never used to exist. The future trend of GANs is expected to impact interesting topics such as image-to-image translation, text-to-image translation, photo blending, clothing translation, and 3D object generation,” he says.
All in all, it’s all change.
“Digital technology is reshaping the world around us, and we want to make sure that this future is inclusive so that everyone can benefit from the changes digital technology brings.
“Mature commercial applications of new technologies – particularly 5G, cloud computing, AI, and blockchain – are speeding up the digital transformation of all industries. This will present enormous opportunities for new industries to be created, which will bring about new roles that we never had before,” he says.