Asking if artificial intelligence is a friend or a foe is to miss the point entirely. While the question acknowledges its inherent power and potential, boiling AI down to this overly simplistic dichotomy ignores the puppet masters – us.
In March a letter was written by the Future of Life Institute, a non-profit headed by the MIT physics professor Max Tegmark, seeking a six-month pause on “giant AI experiments”.
“AI research and development should be refocused on making today’s powerful, state-of-the-art systems more accurate, safe, interpretable, transparent, robust, aligned, trustworthy and loyal,” the letter stated. Elon Musk, Apple’s Steve Wozniak, and 33,000 academics, authors and entrepreneurs signed the missive.
Yet no machine has definitively passed the Turing Test, which says true AI is where it is impossible for a human to tell the difference between human and machine-generated responses to a series of questions. Chat GPT, which blew everyone away when it was launched last year, can still be easily duped into answering nonsensical questions.
But when we consider how digitalisation has transformed the most banal of everyday tasks, what changes will AI bring about in our personal lives and business activities? And should we really be afraid of its potential to one day supersede human decision-making capabilities?
Expectations of and enthusiasm for AI have risen and fallen over the decades, with periods of disillusionment in the industry, characterised as “AI winters”. But those days are over, says Kieran Towey, applied intelligence lead at KPMG.
“Vast increases in computer power and affordability, bigger chips, the advent of deep learning, new generations in the ways that AI is architected to learn – GPT being the most famous – all contribute to where we find ourselves now, on the cusp of the most transformative period in human history, all based on the increasing sophistication of AI,” says Towey.
“While not perfect, the recent upsurge in AI capability is a step change and will dramatically impact businesses but there will be more and more focus on AI controls to harness the power while mediating against its acknowledged limitations.”
Incredible advances
When we consider the speed at which AI is evolving, the plea for a pause makes sense; Towey notes that the computing power used by the cutting-edge AI models doubles every six months. “That’s a 1,000 per cent increase over five years,” he says. “You don’t need to be a data scientist to realise that this increase could lead to incredible advances in that period. If we just look at where we are now, Chat GPT 4, just released, has made significant advances on GPT 3.5 and outperforms the vast majority of humans on standardised testing, achieving the high 90s percentiles on the US SAT tests.”
And there is another step change coming that will necessitate some difficult discussions and forward planning, he says. “Previously, we spoke about automating more blue-collar work, such as transportation and logistics, but now we are very clearly talking about automating white-collar tasks,” says Towey.
Generative AI, a type of artificial intelligence system capable of generating text, images or other media in response to prompts, is markedly different from what we have seen until relatively recently, says Luke McDonnell, senior corporate communications manager at Huawei Ireland.
“What has underpinned this sophistication is machine learning (ML) which is the use of data and algorithms to understand the relationships among data and information, mimicking the way that humans learn and continually improving its accuracy,” he explains.
AI’s ability to lend value to human endeavours is and will continue to be one of the driving forces of the 21st century
— Colin Melody, Deloitte
Through ML, AI-systems are being fed large data sets and mathematical models, and are constantly learning and improving to become more accurate in their predictions and answers. For example, McDonnell points out that customer service chatbots, previously “infuriating”, have become more efficient, more intuitive, and are much faster at giving responses.
According to Colin Melody, director, artificial intelligence and data, with Deloitte, recent advances such as the adoption of cloud-based architecture, particularly in industry and enterprise, has provided the framework for the cost-effective and simplified introduction of AI solutions worldwide.
“AI’s ability to lend value to human endeavours is and will continue to be one of the driving forces of the 21st century,” he says. “AI will increasingly become integrated and adopted into everyday life, allowing greater and broader exploration of ideas and information. Generative AI is rapidly developing and evolving with no end in sight.”
A new skill set
By using generative AI, businesses can gain valuable insights from large amounts of data, allowing them to make more informed decisions, adds McDonnell. “One of the most common debates around the use of AI in business is that it will take jobs. However, in reality, it will open up a range of new positions and careers requiring a new skill set,” he says. “Roles like prompt engineers, who design and create prompts for AI – essentially someone who can tell AI what to research and gets the right answers; or AI ethicists who will be responsible for reducing bias in AI systems and ensuring those systems are used in a fair and ethical way.”
These advances do not just have a commercial bent and in many instances are helping with equality and inclusion. For example, Huawei has developed StorySign an AI-powered app that translates popular children’s books into sign language, allowing deaf children to “read”.
Citing the Future of Life letter, McDonnell says it’s clear that we need to reconsider our relationship with technology and begin playing catch up. “International collaboration in generating AI guidelines and regulations is now more than ever becoming necessary to address the challenges presented by AI,” he says. “Collaboration between industry, academia and government will be important to help develop best practices and standards to maximise the potential of AI and mitigate its potential risks.”
Towey agrees. “We will have to adapt both economically and socially to incorporate AI into our working lives and demonstrate how we can work as ‘augmented humans’, humans working with machines to bring the best of both to be as productive as we can.”
‘We all need basic knowledge’
Melody offers some perspective. “AI is still limited by being primarily single-purpose, focused tools and there is a long road ahead before all its potential utilities can be realised,” he says.
“On a personal level and looking into the future, how we communicate with one another, do our jobs, train up on skills, purchase goods, experience customer service and the other everyday tasks is yet to be seen. Businesses and organisations may also look to further redefine career paths and roles, complementing the efficiencies and benefits of these new technologies with the human skills and expertise to truly realise value.”
Even the most tech-savvy of us may feel lost when it comes to AI and machine learning. It’s worth heeding McDonnell when he asserts that we will all need some basic knowledge of AI. “It is the future and we should understand it, use it and be aware of how it can enhance our everyday lives.”
A useful starting point, he says, is the free online course Elements of AI hosted by University College Cork, which teamed up with the University of Helsinki to bring AI “into the sittingrooms and kitchens of Irish homes”.
“I would encourage those with only even a fleeting interest in AI to give it a go.”