At present, all the world’s data can be stored on a diamond wafer about 2”x 2” in size (Chris Mellor, Storing 25 Exabytes on a Two-Inch Kenzan Diamond Wafer, Blocks and Files, May 3, 2022). Because LLMs require so much data to train and validate, Gen AI-generated data is predicted to grow by 36.5% CAGR through 2030 (Grand View Research). With the unprecedented data captured increasingly by IoT, unimaginable amounts of data on the order of 26.4% CAGR will be created by 2029 (Zippia). This does not account for the increasing amounts of video data from surveillance, satellite and traffic monitoring. Add to those amounts data in company legacy systems and you are looking at amounts that are immeasurable.

How do these trends affect intelligent information management, which is largely comprised of document, content and process management? AI will be a re-creation of reality and a national cultural phenomenon the likes of which we have not seen since the advent of the Internet. However, AI will far surpass the implications of even the Net. There will be no aspect of reality it will not change.

Its Orwellian downside will be at least this: it will be increasingly difficult to distinguish between reality and unreality. Evidence of this already exists with deep fakes, misinformation and disinformation on social media. Its upside, though, will be a radical widespread automation of many operations in every vertical market. What’s more, when you can access news, weather, books, entertainment, health care, etc. via your ear buds, phone and wearables, imagine how TV, radio, weather services, publishing, education, work, healthcare and more will be disrupted or outmoded.

Discussing the specifics of changes in all these areas is beyond the scope of this discussion, but as an example, consider healthcare.

Medical data doubles every 73 days (concultqd.clevelandclinic.org). Increasingly, wearables that monitor vital signs, blood sugar, blood pressure, etc. are contributing to this total – from 2017 with 11 exabytes per month to 2020 with 77 exabytes (Statista). Stored in an electronic health record (EHR), it becomes a dynamic digital twin of a patient’s health condition over time. A digital twin in this case is virtual replica that captures and reflects a patient’s behavior and dynamic health condition. As volume increases, big data analytics increases its value. Indicating illness and correlating intervention with outcomes is the great value of this volume. Some advanced health care systems are getting as many as 70 data streams from remote patient monitoring (RPM) integrated with IoT (Telehelath.HHS.gov). This bodes a paradigm shift in healthcare from a patient/provider relationship to a consumer/device one.

The patient changes behavior — say, exercising more, taking medicine as prescribed — according to what the wearable tells — say, steps a patient takes per day to time they look medication last. This, however, is a relatively simple data analytics solution.

As the authors of the new book, Gigatrends (Thomas Koulopoulos, Nathaniel Palmer, Gigatrends, Post Hill Press, New York, 2024) propose in this section, when that solution is augmented with predictive analytics patients, and more so doctors, can predict the intervention required to prevent a bad outcome — say, weight gain and coronary problems.

In the US, the sickest 5% of the population accounts for 50% of healthcare spending. That percentage skews dramatically towards the elderly. One indicator is telling about population. One in three seniors report falling each year. According to the CDC, the medical cost of fall injuries in this set amounts to an amazing $50,000,000,000 yearly.

Using machine vision, AI and predictive analytics at home, seniors can prevent falls, radically cutting the annual cost of falls — the fastest-growing segment of Medicare. Think about what one application like Zoom did to improve convenience of doctor’s visits. Now imagine the megatrend that is happening from in-hospital care to self- and home-based care, as well as from episodic to preventive healthcare. In the future, hospitals will also outsource most services save ER and big-machine services like radiology-related ones. The cost of genomic testing is also dropping to $100 in the next decade. This factor will permit hyper-personalized healthcare right from birth that predicts treatments from likely illnesses determined by a person’s genetic make-up.

The authors propose an imaginative solution for managing continuity and correlation of health care they call Digital Healthcare Advocates. These AI devices autonomously manage continuity of a patient’s healthcare information across all devices like EHRs and data repositories; coordinated care across PCPs, specialists, hospitals, insurers, pharmacies, etc.; predict problems; and monitor and reinforce patients’ behaviors to comply with treatments. In the US, the cost of noncompliance with medical treatments alone is estimated as high $290 billion annually in more doctor’s visits, hospitalizations and treatments. The cost of ER visits in 2017 was $76.3 billion (census.gov), and hospitals had to write off $42 billion in 2022 in unpaid for ER care (aha.org). With digital advocates, patients would interact with them as AI software solutions. Such an inventive solution would advance a national cultural phenomenon that would improve service and cut costs for all concerned.

A wholesale overhaul of the US healthcare system is a daunting venture. However, facts bear out its necessity. By 2040, the authors cite federal government spending on healthcare is estimated to total 40% of US GDP; by 2060 it’s predicted to be 60%. This is unsustainable.

Increasing peoples’ quantity of life is a costly undertaking with diminishing value over time. Increasing their quality of life makes for a healthier, happier, more productive population that, rather than being a burden on, contributes longer to the economy. AI radically simplifies reality for medical personnel — doctors interact with EHRs, patients with wearables and digital advocates. This cuts their time with tech rather than addicts them to it.

A year ago at a trade show, document/content/KM vendors were chasing the low-fruit AI use case of customer service, especially in contact centers. AI quickly pulls up the right answer to a customer query when the CSR prompts the LLM populated with relevant data, thus expediting customer service.

How much of a leap would it be to impound patients’ healthcare data augmented with similar patients’ data in an LLM (as in the CSR’s LLM) so it functions as a digital advocate (the patient would prompt the LLM like the CSR). Patients might even prompt the advocate from an interactive wearable and access information about the patient’s condition like diabetes, independent of the doctor, and treat themselves. Fewer doctor interventions cut patient costs and reduce doctor burnout, a serious problem. More problems like falls, high blood pressure and sedentariness in the elderly could be patient-monitored and managed and radically reduce the cost to the government. AI creates new ways of thinking that open up new solutions. Measureless data and improving Gen AI is what IIM has now. Why can’t visionary vendors explore the uncertainty inherent in a quantum computing world we live in now with hope for a better future and achieve what seems impossible now but the closer they get to it becomes all but certain?