Mention artificial intelligence (AI) and the reaction varies from blank to reluctant to frightened to downright dismissive. AI enthusiasts are mostly Tech geeks and IT companies who see a creative and financial bonanza round the corner. This is normal. This is understandable. But, this is a bit uncalled-for too. Every new tech invention is supposed to be an existential threat for industries, businesses, jobs.
The VCR was supposed to kill the cinema industry, the internet was supposed to annihilate TVs, the email was a guzzler of courier services, etc. None of that happened. Yes, businesses were disrupted. Yes, people were dislocated. Yes systems were rendered obsolete. But they gave way to better, faster and wider applications and usages that benefited multi sectors at large.
The S curve for technology adoption explains this non-linear acceptance of an invention. Innovators, early adopters, early majority, late majority and laggards appear along a slow beginning, sharp growth, a quick flattening and a steep decline.
The AI curve is still in the innovators’ stage where hardly 2.5% are really understanding and willing to use it in their personal and professional lives. The jury is still not out on whether it is going to let us control our lives more efficiently or it will start controlling our lives.
It may be too early to give it an overall green tick but one industry that is definitely going to benefit radically due to AI’s enablement power is healthcare. Let us look at the dramatic change that can take place if AI is used proactively in the healthcare industry:
Transforming data, diagnosis and detection— AI is a super genius at assimilating tons of data and coming up with accurate reports in almost no time. Using this feature the healthcare industry can improve its timeliness, accuracy and delivery multiple times.
A study carried out by National Health Service Britain, showed how usage of AI is really helping it to overcome its issues of long service times and delayed patient care. AI is used in every facet of healthcare, with 34% of NHS use cases being diagnostic. This includes AI-driven image analysis, pathology, and endoscopy. In future AI is going to play a major role in identifying illness and disease in Britain.
The second most common application of AI is automation/service efficiency, followed by population health. The latter is focused on ‘leveraging the strength of AI when dealing with large patient datasets to help predict and prevent diseases in the general population’. This could be huge in terms of public health diseases and pandemics like COVID.
Reducing research and development costs— Developing a drug can take 3 to 12 years. Some drugs cost $ 2 billion, some more to develop. Imagine the issues of a 12-year changing environment. AI has and will have a revolutionary impact on this long cycle, especially data analysis.
The early phase of drug discovery is a tireless process of reading and analyzing scientific literature and calculating and testing drug interactions. AI can automate much of this and save billions of dollars in the process. In turn, this can drive down the end cost of drugs too.
Funding for AI drug discovery rose 3,800% between 2016 and 2021.That just shows where the expected returns are. According to Financial Times, Moderna used an AI model trained on 20,000 unique mRNA sequences to help design and manufacture its first batch of COVID-19 vaccine for testing in just 42 days. AI shows great promise in speeding up the drug discovery and clinical trial process.
Safeguarding human surgery errors— Medical errors are killers. They can maim, cripple, destroy and stunt human health. According to many studies one in seven medicare patients face a medical mistake. In Pakistan this may be much higher.
Every day there are scams of doctor’s neglect, fake tests and wrong diagnosis, causing misery to hundreds of families. The tragic 200 patients’ death and over one thousand sick at Punjab Institute of Cardiology in 2012 was just one such scam that exposed the fatal human errors.
In the UK NHS study AI robots for surgery is valued the highest ($40 billion by 2026). Virtual nurse assistants were ranked second ($20 billion), which will free up nurse workload and counter labor shortages. One of the main problems of erroneous treatment is erroneous diagnosis.
AI sharply reduced the margin of error. Currently, 86% of NHS stroke treatment facilities are using AI diagnosis tools to speed up detection and treatment. This will be 100% by the end of the year. This is part of a £21 million AI diagnostic fund that also includes AI technology for chest X-rays to detect lung cancer.
The global race for AI is on. The local face of AI is off and on. Pakistan’s healthcare industry is struggling with lack of resources, skills and technology. While everybody recognizes the usage of technology and especially AI, few have embraced its potential. Most tech universities are now having some basic AI centers.
NUST and COMSTECH have done good work on it. There is an excellent institute PIAIC, i.e., Presidential initiative on Artificial intelligence and computing launched in December 2018 that aims at revolutionizing education and industry through education and technology. Another initiative is NIAIS, i.e., National Initiative for AI and Security established in 2019 that aims at security areas like cybersecurity, etc.
What is needed is a more integrated and holistic approach to form policy and strategy at national and industrial level. These initiatives need to be synergized to create a proper think tank for policy planning and an execution body that ensures implementation. A 5- point agenda is required to develop a comprehensive AI development impetus:
Form an Apex AIC (Artificial Intelligence Council). This should have the highest representation from Ministry of Science and Technology, Ministry of IT, Ministry of Health, Ministry of Education and Ministry of Defence. This council should make an AI adoption plan till 2030.
The AIC should make public-private partnership a model. Associations like PASHA, PPMA, etc., should be part of co-development of the strategic, investment and operational side of AIC. All sectors of healthcare should be represented to create the holistic approach.
Use NAVTEC/TEVTA, etc., for Tech Training-Create policies of having at least 10% of tech training in these institutions on AI programs. Develop affiliations with international universities, etc., to develop online programmes for masses.
Encourage AI competition in students— During Covid it was NUST students who developed prototypes of the much needed ventilators. Time to engage incubator project competitions amongst students in AI development. Winning students to be sponsored for foreign education, etc.
Legislate on mandatory AI education-Pass bills for introducing AI technology as part of medical education. CPSP credits to be given to medical students training on AI surgeries etc.
AI is not just a keep up game, it is a game changer. Pakistan with high unemployment can really take this as an opportunity to provide employable skills to the world. Budgets are stringent but existing resources can be sharpened to provide start-up platforms. It is a race not just against machines but against life. Time for a little more care for healthcare.
Copyright Business Recorder, 2023