Growth. Intelligence, and learning Can machines last longer than us now? From our morning emails, and the elation of online shopping to cars that operate on their own, technology is sometimes more perceptive than we are. Artificial intelligence is practically a juggernaut, making its way into practically every sector in the world, with far improved efficiency.
The skepticism around artificial intelligence leads to the question: Can they possess values like honesty, integrity, compassion and empathy that are instilled in us at a young age? In relation to Lethal Autonomous Weapons System (LAWS), technology expert Peter Ansaro emphasised on the fact that choosing to pull the trigger of a gun and putting that weapon into use is a moral act- an act that requires complete cognizance, sensitivity and a sense of responsibility. Will sentinel guns have these? No, of course not. Based on quantitative data, a machine is made to think. Skills like righteousness and conscientiousness, ethics and morality, can’t really be stringed into an algorithm. How can humans program machines with these qualities?
Firstly, to make AI more human-centered, it is imperative to eliminate any sign of AI bias. AI bias is caused by cumulative human biases that are passed into the AI systems created by humans. These can be biases based on colour, race and gender. In simple words, will black lives, abortion rights and profound issues like these matter to a RoboCop? This is further articulated by the artificial intelligence tool in the USA courtrooms called COMPAS, which was designed to predict future crimes. Past criminal records, employment statuses and age are also taken into account and Black defendants have been labelled as “high-risk” several times.
Let’s talk about something that is imminently going to affect us- the Metaverse. We all know that this is going to be our imminent future, strongly based on AI. But we fail to realise that quite egregiously, in the long run, a concept like this would eventually foment an atmosphere of melancholy, desolation, and sequestration amongst the users as they would be trapped within their own ‘meta-world’, with zero to no human interaction. Another salient factor is that one would eventually begin to experience nuances of frustration, remorse and clinical depression with hardly a tinge of delight.
Making AI more empathetic starts with having a diverse team of workers and programmers, holding consultations with various stakeholders, including people from vulnerable populations, before building a data set.
Healthcare is another field that is rapidly advancing in AI. Adherence of treatment and engagement of patients has greatly impacted the outcome of the health and well-being of the patient. If patients are not actively participating in their treatment and are unable to interact with their doctors, the probability of them to come for follow-up appointments or implement necessary changes like diet or exercise decreases. With many people lacking a connection in the world today, AI makes enlarges this gap. In a survey consisting of over 300 clinical professionals and healthcare executives, 42% of participators said that fewer than 25% of their patients were highly engaged. To not only develop better patient outcomes, but also increase engagement, AI models can be programmed to send message alerts to patients’ phones, to take medicine or to go out for a run. These messages need not be mundane or robotic, they can be encouraging and motivational, based on the personal choices of the patient. This will also help the system establish a large database based on the likes and dislikes of patients, enabling them to cater to forthcoming patients more effectively. This can not only put importance on physical health but also boost mental health.
AI in the field of surgery and medicine truly has a long way to go. Did you know there have been proved instances of AI having developed a racial bias? In 2019, a group of researchers discovered a defect in an algorithm used in hospitals in the United States on more than 200,000,000 people. Abysmally, this algorithm favoured white patients and gave them priority over black patients when it came to determining which patient would require extra medical care.
Now, the basis of distinction was not directly colour, but when patterns and statistics were analysed, the truth overshadowed the false picture that was painted. It was all based on cost- on average, black patients had lower healthcare costs in comparison to white patients, due to which the AI was tuned a certain way to determine that black patients would not ‘need’ additional healthcare. Unfortunately, whenever humans are involved, they play the role of the devil’s advocate and increase the risk of hidden biases to slip into the AI’s algorithm. Abating these inadequacies would obviously improve AI and prove to be more beneficial to humankind.
AI still has a long distance to cover to reach empathy. It poses several risks such as job automation, leading to unemployment, malicious use of data, privacy breaches, widening socioeconomic inequality, autonomous weapons potentially leading to an AI arms race and even stock market volatility.
Ironically, AI is also going to be what helps us resolve these facets, when and if used appropriately.
Despite these drawbacks, AI can revolutionise the field of surgery in the following four ways- preoperative planning, intraoperative assistance, early diagnosis and AI-enabled remote surgery. AI analyses and learns from large sets of data, recognizes new trends, relieves cognitive and physical stress and also introduces novel methodologies to better medicinal practice all across the globe.
From quickly analysing tremendous amounts of clinical documentation to scanning radiological images and mammograms for early detection of diseases based on multimodal information – AI does it all. By leveraging AI, healthcare systems can become smarter, faster and way more efficient, providing care to millions of people worldwide. The global market for robot-assisted surgeries is projected to increase from $4B to $18B by 2030.
In terms of recent technological advancements, Amiko Respiro enables datadriven respiratory treatments, catered to your personal needs and conditions.
One in 12 people suffer from asthma, and Respiro makes living with this incurable condition much easier using artificial intelligence. This smart device basically fits on inhalers and keeps tabs on your doses and records so you don’t have to. Children as young as 3 and adults as old as 90 suffer from asthma. For them, it becomes really tedious to keep a track of their doses. Amiko came up with a one-time solution for this. Respiro is a ‘smart’ inhaler embedded with machine learning, advanced sensor technologies and AI powered digital health solutions which can assist healthcare professionals everywhere.
Another breakthrough in the surgical field is the Da Vinci Surgical System (no, the painter has nothing to do with this), which gives a surgeon an advanced set of instruments that are used in performing robot-assisted surgery, an advantage of which is that it is minimally invasive. Now, the robot doesn’t actually “perform” surgery. It functions according to the commands given by the surgeon via a console. This system enables surgeons to operate through bare minimum incisions and also provides highly magnified, 3D high-definition views of the surgical area.
At the same time, AI can be used for avoiding surgery too. Some diseases, like the Peripheral Artery Disease can be treated without surgery if detected early on. In late detection of PAD, however, the procedure can be as extreme as surgical amputation of the limb. The National Health Service of UK has also initiated a program called Accelerating Detection of Disease which will use AI to develop solutions for the early diagnosis and anticipation of illnesses. Robotenabled surgeries have shorter recovery periods, lower post-surgery pain and also fewer complications. AI and computer vision is furthering helping surgeons improve spinal surgery, which is an extremely complicated and high-risk procedure. Augmented reality improves surgeons’ intraoperative vision and enables them to easily access the desired area. Tissue tracking assists with differentiating between organs, while endoscopic navigation improves techniques to maneuver toward a targeted location.
In conclusion, developers do need to re-strategize, rethink and re-assess algorithmic concepts and program systems related to artificial intelligence and arrive at a reasonable consensus. We live in a real world with real people and real problems. These problems need to be contemplated upon and action needs to be taken. If not now, when? If not us, who? AI is the future. It’s about time we realise that and make necessary improvements.
Ishana Sharma, Grade 10, Cathedral and John Connon School, Mumbai.