The momentum on AI continues to accelerate and everyone from Elon Musk with Steve Wozniak (Apple Co-Founder), to Bill Gates (see last blog) has an opinion. Some predictions are for a catastrophic future as predicted by many science fiction movies from The Matrix (1999) to Transcendence (2014). Many others forecast a time when AI will do many routine things that we humans find boring or repetitive. The last I heard is that AI will eliminate the need for Auditors, Bookkeepers, financial advisors, underwriters, economist, legal assistants, librarians, online teaches, real estate agents, recruiters, journalist, copywriters, computer coders, software engineers, data analyst and programmers, just to name a few. As I write this, I am scheduled to meet with a group that has suggested they can help us create a new software program using NLP (Natural Language Processing) in a fraction of the time and cost of traditional creation. I am fascinated, but skeptical as I’m not sure we’re there yet.
The headline reads: As Elon Musk Calls for A ‘Pause’ On AI Development, Will The AI Wars Reach A Temporary Truce?
Is this irrational fear, hyperbole, or just a contrarian view? Mr. Musk, clearly one of the most brilliant people of our generation, appears to be genuinely concerned. In a letter Mr. Musk calls for developers to “pause” the training of advance AI systems, until a “regulatory framework can be established”; however, while bureaucrats are happily rubbing their hands together at the thought of more regulation, the questions being posed are fundamental. Regulate what? Regulate how?
What appears to be unfolding is a replay of the Old West Gold Rush days. Only this time, the rush is coming from the largest, most powerful, and most advanced technology companies to get to the lead in the development of AI. What does “lead” mean as AI is, by definition, always changing and evolving?
Today, this advancement is not being led by universities, but by companies with nearly unlimited resources like Microsoft and Google. The rush is not only in the USA, but also in many countries like China, India, and others. Top AI chip maker, Nvidia, is taking off with the development of more advanced chips, some costing as much as US$10,000 EACH.
We are confident that AI applications in healthcare will be lifesaving. As mentioned in the last blog, there are half a dozen applications that we can use AI for today. Since the wide-spread use of Electronic Medical Records (EMRs) in the USA, with the mandates created, funded, and enforced by 2010 laws, we have nearly full penetration of the patient medical record data in the USA.
I asked ChatGPT this question earlier: Have Electronic Healthcare Records in the USA helped save lives? Do we have any statistics? Below is the reply.
Electronic healthcare records (EHRs) have the potential to improve healthcare outcomes and save lives by providing healthcare providers with more complete and accurate information about their patients. However, the impact of EHRs on patient outcomes and mortality rates is complex and difficult to measure.
There have been some studies that suggest that EHRs have contributed to improved healthcare outcomes and reduced mortality rates. For example, a study published in the Journal of the American Medical Association (JAMA) in 2018 found that the use of EHRs was associated with a lower 30-day mortality rate for patients with acute myocardial infarction (heart attack) compared to hospitals that did not use EHRs.
Another study published in Health Affairs in 2019 found that hospitals that adopted EHRs saw a significant reduction in mortality rates for patients with certain conditions, including pneumonia and congestive heart failure.
However, there have also been concerns raised about the potential for EHRs to introduce new risks and errors into the healthcare system, such as medication errors or data entry mistakes. It’s important to note that the impact of EHRs on patient outcomes will likely depend on a number of factors, including how the EHR system is designed, implemented, and used.
As ChatGPT stated, overall, while there is some evidence to suggest that EHRs may have contributed to improved healthcare outcomes and saved lives, the relationship between EHRs and patient mortality rates is complex and further research is needed to fully understand the impact of EHRs on healthcare outcomes. Not the unequivocal statement one might hope for from the computer, but the prompt engineering could be a little tighter as well.
I am smiling as I read the final paragraph, I have underlined what I believe is the key statement. I now believe that after US$100s of billions of dollars invested in EHR (in USA alone), we have incredible amounts of data in now deeper and wider silos. Just maybe, that is the monumental opportunity for AI to process this data, that no modern EHR can do. Maybe, just maybe, we are now at the foundation to “advance the science” ™ one of the stated outcomes of our Company’s mission. We can do better than to just dig through digital and scanned patient records to aggregate, synchronize, and analyze a patients’ past records to enable a more accurate forecast of their future health. We think the opportunity for AI is stunning in size and impact on society.
Stay tuned …. There’s more to follow.
Noel J. Guillama, Chairman