The sci-fi thriller “M3GAN” depicts an AI-driven, lifelike android, who commits unspeakable atrocities in fulfilling its mission of becoming a friend and companion to a little girl. The film has a storyline that echoes all-to-common misconceptions of AI; such dramatic speculation has resulted in a throng of arm-chair Luddite commentators issuing dire warnings about our future AI-driven overlords surpassing human intelligence and taking over the world.
The reality of AI differs from the dystopian imaginations of Hollywood writers. While artificial intelligence and college campus thought experiments around “Singularity” are popular, AI is not, in fact, going to become our new master. What then, will AI accomplish? The most effective applications are not in the development of lifelike robots, but rather, in decision support, enhanced decision-making, business transformation, and systems modernization. In areas as diverse as healthcare and insurance, finance, and manufacturing, AI has already gained a foothold, delivering unmistakable benefits, shareholder value, and convenience for consumers.
A 2022 PwC AI business survey1 indicates that companies with proof-of-concept AI projects are ready to scale and already have active use cases, with many examples already showing significant ROI in the areas of business transformation, enhanced decision-making, and modernized systems and processes, as well as in overcoming labor shortages.
AI in healthcare: The doctor is in
Novel innovations are slow to take hold in healthcare compared to any other industry, and with good reason. With concerns over compliance, privacy, and repeatable accuracy, the industry is heavily regulated and must proceed with caution. But when AI-based benefits do accrue, they can be transformative and even life-saving.
We are not likely to see Star Trek-like holographic doctors any time in the near future, but the use of AI in healthcare has already led to important advances in diagnostics, accurately detecting irregular test results, and even assisting surgeons in performing delicate medical procedures. The ability to minimize human errors and life-threatening complications is unmistakable. On the paperwork side (and in healthcare, there is a lot of paperwork), AI automates routine tasks in documentation, data entry, and billing, and excels in clinical decision support, serving as a tool to support, but not replace the skills and knowledge of medical practitioners.
Risks and challenges
AI is not without risk. AI talent is expensive and scarce, although according to the PwC study, many companies are addressing the issue by retraining in-house talent to work with AI and reorganizing teams to make better use of the AI experience, while respondents are also one and a half times more likely to address their AI needs with third-party vendors.
AI, in its current iteration and with tools like ChatGPT providing by far the most human-like level of interaction in the history of computing, also carries the challenge of responsible use. Can AI do bad things? Absolutely, as is the case with any type of computer program. Hackers and other villains have been manipulating technology for years with ill intent, and it’s a given that AI tools like ChatGPT will be no different. The most important imperative in AI usage is to assess each AI model for robustness, bias, and fairness, and to ensure that checks and balances are in place. Further, when used in highly regulated industries like healthcare, compliance with applicable regulations moves to the top of the list.
This article was written by a human person
ChatGPT did not, as some readers may wonder, write this article. Writers, as a group, tend to have overactive imaginations and have recently begun to obsess over whether such AI tools will come for their jobs. It’s not likely to happen. Writers do use a host of tools, which often get replaced and upgraded over time. I began writing using a manual typewriter, which gave way to an electric and then later a PC. We use support tools, references like the Chicago Manual of Style, and search tools like Google. ChatGPT is just the latest iteration of a wave of progressively more sophisticated support tools. Although it will not replace me, ChatGPT does make a fine research assistant, despite its inability to bring me coffee and donuts in the afternoon.
Twelve years ago, Apple gave birth to Siri, giving us a sneak peek at what the future may hold. But to be accurate, Siri is not AI, it’s just a search engine with a friendly-sounding human voice interface that is primarily a command-and-control system. ChatGPT has changed the game with a true AI tool that uses a large language model capable of improvisation and handling complex tasks and Google, Amazon, and Baidu are already in a race to compete. This tool and other AI innovations represent a major step in improving accuracy, compliance, and the end-to-end customer experience.
Reference
AUTHOR
SUBJECT TAGS
#AI
#human intelligence
#business transformation
#healthcare
#chatGPT