Artificial intelligence must work for all of us – not just the rich and powerful
Mike Hedges – Gŵyr Abertawe
We are now in the fourth industrial revolution, the Artificial Intelligence (AI) revolution, building on the ICT revolution of the 1980s. This must work for all of us not just for the rich and powerful and benefit society.
We are seeing fundamental changes taking place in how the global production and supply network operates not only the continual automation of traditional manufacturing and industrial practices, using modern technology, the internet and large-scale machine-to-machine communication, but also in the use of machine learning to perform some technical tasks better than humans.
This then results in improved communication, increased automation, increased self-monitoring, and most importantly the use of smart machines that can analyse and diagnose issues without the need for human intervention.
Machines have for many years improved human efficiency in performing repetitive functions, doing it faster and more consistently than people from spraying cars to auto mated accounts.
The combination of machine learning and computing power allows machines to conduct increasingly complex tasks.
AI has a wide range of applications across all sectors of the economy. It gained prominence following advancements in deep learning during the 2010s coming to public attention when computers started beating grand masters at chess, solving sudoku faster than humans and finally beating expert Go players.
Its impact intensified in the early 2020s with the rise of generative AI, where models can engage in verbal and textual discussions and analyse images.
One of the areas that artificial intelligence can be used to improve productivity, efficiency and outcomes is health.
Health care has become more expensive due to the cost of patient care and drugs. It is an area where AI can help increase efficiency, algorithms can analyse medical images, patient data, and other information to assist in diagnosing diseases, often detecting patterns and correlations that might be missed by humans.
Strokes
AI can help diagnose lung cancer more accurately and predict heart attacks and strokes more effectively than medical professionals.
We have had publicly reported that the application of AI algorithms in areas such as ophthalmology has ensured increased accuracy in the screening and diagnosis of glaucoma and cataracts.
From published research we know that AI can help develop treatment plans tailored to individual patient needs, considering factors like their specific condition, genetic makeup, and other relevant information.
In a study by Yu et al, AI software used quantitative histopathology characteristics collected from 2,186 whole-slide pathology pictures from the Cancer Genome Atlas to discriminate between primary lung adenocarcinoma and squamous cell carcinoma.
In a multicentre, non-interventional trial, including 120 pulmonologists from sixteen hospitals in five different European countries, Topalovic et al, showed that AI performed better than pulmonologists in the interpretation of pulmonary function tests.
Surgeons
Pulmonologists’ pattern identification of pulmonary function tests met the recommendations in 74.4% of instances, and pulmonologists made proper diagnoses in 44.6% of cases, whereas AI precisely matched the PFT pattern interpretations 100% of the time and gave a correct diagnosis in 82% of cases.
We have research telling us that AI-powered devices help surgeons perform minimally invasive procedures with greater precision and accuracy, reducing the risk of errors and complications.
AI is being used during surgery to optimize force, detect positive surgical margins, and even automate specific steps.
AI-powered patient monitoring can monitor patient conditions in real-time, providing alerts to healthcare providers if there are changes that require attention.
While it has been reported that currently autonomous robotic surgery appears to be a long way off, cross-disciplinary collaboration will certainly improve AI’s potential to complement surgical treatment.
Already we have had examples of robotics and neuro-navigation technologies assisting in minimally invasive surgery.
AI can be used for personalized patient communication including sending patients schedule reminders, health tips, and suggested next steps, improving patient engagement and adherence to treatment plans.
AI can also be used to predict which patients are at risk of becoming frequent users of emergency services, allowing for earlier intervention, and potentially reducing demand on emergency departments.
When people contact financial businesses, the first stage is a chat with AI, only being referred to a human operator when AI cannot answer the question being asked.
Wales
Why cannot AI be used to produce surgery lists and, answer questions and to contact patients informing them of their date and time of operation.
The last industrial revolution saw the growth of large companies such as Google, Apple, Microsoft, and Facebook.
The area around Silicon Valley hosted a large number of the major companies created as part of the ICT based revolution.
Our challenge in Wales is to become the home of the major companies in the AI led industrial revolution – starting with health.
Support our Nation today
For the price of a cup of coffee a month you can help us create an independent, not-for-profit, national news service for the people of Wales, by the people of Wales.
Get more trusted Welsh news
Choose Nation.Cymru as a preferred source in Google News to see more of our journalism.
But not simply as a tax convient investment vehicle for giant US/Chinese Mega Corps Data Banks which are energy and water eaters and provide diddly squat jobs, Data Centres are the new form of colonialisation. We neeed that new green energy to support our foundation industry recovery e.g. for Steel Making and Hydrogen production…..
Interesting – but probably not a full-blown industrial revolution just yet. Waiting for robotics to kick in as well as automation.
no it’s definitely full blow industrial revolution. Robotics and AI will be a whole new ball game. LLM is like the automatic loom, AI+Robotics is like having a car.
I read the WG policy and strategy document: “AI Cymru: Shaping a Smarter, Fairer, More Prosperous Wales published 18.11.2025. AI, it states, should be used to “equip people in Wales with the skills to thrive in a world shaped by intelligent technologies”. Indeed, and this should be a top priority of the new WG. The documents itself is rather too full of Blair era management speak and generalities: “A values-led 4 Pillar framework: “Economic Growth, Educating Wales, Equitable Delivery, Excellence & Trust”. Principles to guide AI adoption: “Ethical… Empathetic… Enterprising… Effective” and so on. Bit light on specifics.
It is important to differentiate between different types of “AI”. Machine learning has been around for quite a long time and is very effective, and is very likely the technology being used for the examples in the article. Machine learning requires relatively small amounts of power and does not rely on training using masses of copyrighted materials. Good machine learning models are effectively excellent pattern-matching systems. It has its limitations, but generally speaking works very well when targeted at specific problems. It’s all about the right tool for the job. Generative AI (LLMs – large language models) and the like,… Read more »
Good review/analysis – thank you.
It’s Mike “Treacle” Hedges. He just lists things without any critical engagement or insight. Does it every time.
Generative AI is really just a giant prediction engine. Sure.
Also: The brain is really just electrochemical signaling
Why cannot AI be used to produce surgery lists and, answer questions and to contact patients informing them of their date and time of operation. Because AI is not good enough yet. Medical stuff is a much higher bar to clear. AI is very good, don’t get me wrong but it’s not that good. The ability to answer incorrectly is embedded in the training data. Contacting patients and informing them of date and time of operation is quite automated already. Usually a text message as soon as it’s booked or a letter. It would be much better to start with… Read more »