Artificial Intelligence (AI) is the new “cool kid”

On the off chance you have been living under a rock or living on Mars to get more candy bars (my version of opening with a joke), you might have missed the explosion of artificial intelligence (AI) into our daily lives. It’s everywhere. The news loves to cover the latest version of it almost daily, it seems. Businesses are rapidly implementing it into their business models, and governments are trying to decide how to best regulate it before it’s too late. Newly elected President Elon Musk is trying to keep its development measured and open, while top AI influencers like Sam Altman seem to want to move faster towards commercialization. Despite many great minds working on answers, nobody seems to know for sure what the right answer is regarding AI’s development and commercialization, which is a potential problem in my ever-so-humble opinion.

Much like everyone else, healthcare is also caught up in the excitement and buzz surrounding AI. Depending on the news article or some venture capitalist's profit fever, AI is going to reduce the cost of healthcare to mere pennies while increasing profits to a billion zillion dollars, save billions of trees by making the need to fill out 2,363 forms before a doctor’s visit obsolete, and with a little luck, finally make healthcare realize that the fax machine can finally be retired. I know, that last one might be reaching for the stars, but a boy can dream, right?

Time is the enemy

Excitement is good, innovation is often beneficial, and healthcare needs both right now. I don’t think anyone in healthcare would argue against this. However, there is one aspect of AI in healthcare that I don’t believe is receiving nearly enough attention. How does a patient utilize AI? More specifically, how can a patient utilize AI when so much of their health care is performed because time is the enemy?

For example, as someone who fell off a cliff while rock climbing, I’m obviously not going to break out my laptop to ask AI to provide me with a comprehensive compare/contrast report of all local hospitals within Boise. As a rock climber, I’m obviously not going to spend time asking AI to provide reports on all the emergency rooms of local hospitals on the off chance I do fall off a cliff. Right or wrong, health care is often a reactionary process whereas AI is a proactive tool. In this case time meant pain so I’m going to go to the closest emergency room available to where I was climbing that night my left arm was in pieces and my head was gushing blood. Innovative tools are only as good as the time available to take advantage of them.

Since I’m currently waiting on Fred Meyer to fill one of my heart medications, I thought I’d try to create a simple, beginner’s “how-to guide” for employing AI when time is not a factor. If we can become familiar with AI in our free time as patients, this new innovative tool will become easier to use when time is a factor, I must believe. As all of us chronic patients know from too much experience, our lives are hard enough, so we don’t need a tool that is advertised to make our life easier but, in fact, complicates it even further.

What’s up AI, you friendly or a headache ready to happen

For this guide, I will be using ChatGPT. There are many different AI platforms out there a patient could use. Some are free, others not so much. I’m using ChatGPT since it was one of the first major platforms that received widespread attention from the press, which means its policy towards data has been scrutinized. As a patient, I would encourage others to experiment with several different platforms so you can receive the best possible information. Now, I would recommend that no one use Twitter’s AI, which is called Grok, because of data privacy concerns.

Quick adventure into politics and health data. With the new administration, data privacy is going to be more important than ever. There are already conservative Project 2025 types who are advocating using period trackers to make sure pregnant women conform to their religious beliefs. Others are advocating for setting up checkpoints throughout towns to monitor pregnant women, and who knows what RFK Jr. and his weirdness will want to do. Voluntarily giving Grok and Musk private healthcare information could end with you being punished for receiving lifesaving care if it violates Project 2025’s religious beliefs. Sharing can be dangerous to one’s health right now because AI is still very much the wild west, especially with partisan extremists like Musk and Trump.

Enough politics let’s get back to creating a “how to” guide. For information, I will be sharing my last three MRI reports with ChatGPT regarding my cervical spine. I originally wanted to share my cervical spine X-rays with ChatGPT with the hope it could create a better comparison and contrast report than I received from the radiologist. But when I asked if it could accomplish this task, ChatGPT said no. Also, the free version can only examine one upload at a time, it said, so I had to copy and paste the three MRI reports separately.

Let’s take this thing out for a spin

Before I begin, I want to give you a piece of advice that has served me well over the years. With any new technology like AI, it is almost impossible for someone like me (not a tech expert, only an enthusiast) to break it. I’m not going to ask it a question or do something that will make its head explode. The worst that can happen is I ask it to do something that it can’t do or understand, which it will tell me. When this happens, that doesn’t mean the tech is bad; it means I need to modify my request and try again. AI is a learning process; it’s a learning-by-doing self-advocacy program.

Can you create a summary of all three findings for me? After uploading my three MRI reports regarding my cervical spine, I asked the AI if it “can you create a summary of all three findings for me? Probably not the best-worded question, but the platform responded by saying, “I can create a consolidated summary highlighting the progression, improvements, or new findings across all three.” Since I’m trying to create an introductory type of guide for using AI, I did not want to make my initial request complicated or detailed. I’ve heard from other patients that ChatGPT can create reports in basically any format that you might need. Like I mentioned earlier, you are not going to break the AI, so play around with different types of reports that would be useful for you.

Since I did not know how I wanted the AI to present the information to me, it came back with a basic bullet point style report. The language it used was very similar to the radiologist’s report, meaning there was a fair number of medical terms and jargon. That said, each bullet point was short and to the point, I felt like. Here is a small example (if you have any questions about the medical terms or jargon, please ask. My goal for posting the result was to give you a visual example of the final product that the AI produced and not so you could discover more about my spine issues):

Cervical Disc Levels:

• C2-C3:

• Facet arthropathy with mild-to-moderate neural foraminal stenosis bilaterally.

• C3-C4:

• Minimal disc bulge and vertebral spur with left facet hypertrophy.

• No significant spinal canal or neural foraminal narrowing.

• C4-C5 (Fused):

• Post-surgical changes with vertebral spur and mild left-sided facet hypertrophy.

• Mild neural foraminal stenosis.

As we all know, there is more than one way to tell a story. This is a particularly useful skill to have while at a doctor’s appointment. With that in mind, my next question for ChatGPT was, “Just out of curiosity, what other formats could you present this information in?” ChatGPT responded by stating it “can present the information in various formats depending on what suits your needs best.” Here are some options it presented to me:

1. Tabular Format: A table summarizing findings for each cervical level across all three reports.

For some odd reason, squarespace would not let me copy and paste the chart ChatGPT produced for me. I can tell you that the large was basic, but well done in my opinion. The columns and rows made sense and the content was well worded. I sincerely apologize for not being able to show you the final product that Chat created for me.

2. Visual Chart or Graph: I can describe a way to visualize the progression, such as:

• Bar Chart: Levels of stenosis severity over time.

• Timeline: Showing major findings and changes across reports.

  3. Bullet Points by Theme: Grouped findings based on key categories:

• Stable Findings: Fusion sites at C4-C6 remain intact.

• Progression: Notable worsening at C7-T1 over time.

• Unchanged Findings: No significant spinal canal stenosis across all levels.

  4. Narrative Comparison: A chronological story-like format highlighting changes and stability over time, useful for detailed review with your doctor.

  5. Infographic: A visually engaging layout with icons and color-coding (e.g., green for "no change," red for "progression"). This could require external tools to create, but I can guide you on the content design.

As a patient advocate who utilizes blogging as a tool for my advocacy, I was intrigued by the “Narrative Comparison” option. Often, pain clouds my chronological life, which means certain aspects of my symptoms and life are probably being weighted incorrectly when I present my story to my doctor and you the reader, so this option has a lot of appeal. Also, a narrative comparison approach could be an intriguing format for me to use when I share my story with you. To explore and learn more about this type of report, I asked ChatGPT to provide an example of a “narrative comparison by using my MRI reports as information.” This is the product it created for me:

Narrative Comparison of Cervical Spine MRI Reports (2020–2024)

In April 2020, your cervical spine MRI revealed the effects of a previous surgery: an anterior cervical discectomy and fusion (ACDF) at the C4-C6 levels. The fusion was stable, with no signs of hardware complications. At that time, mild degenerative changes were noted, particularly at C2-C3, where mild-to-moderate neural foraminal narrowing was observed. Mild stenosis was also seen at C6-C7 and C7-T1, with minimal facet joint arthropathy at other levels.

Two years later, in February 2022, the MRI findings showed some mild progression. The fusion site remained stable, with no significant complications. However, at the C4-C5 level, foraminal stenosis had progressed from mild to moderate, and the left-sided narrowing at C6-C7 had worsened slightly to mild-to-moderate severity. Degenerative changes, such as facet hypertrophy, remained consistent across levels, with no significant changes to the spinal cord or overall alignment...

Summary of Changes Over Time:

• Surgical Fusion (C4-C6): Stable and intact across all reports.

• C2-C3 and C3-C4: Degenerative changes (spondylosis) have remained mild, with no significant stenosis in the most recent MRI.

This is not the entire report that I received. Due to its length along with the overall length of this post, I’m only sharing a portion of the report I received from ChatGPT. That said, I think you get a clear picture of what the platform can produce if you request a Narrative Comparison format.

Conclusion with a Big Al Ted Talk flare

Data, data, data. Despite the thrill and excitement surrounding artificial intelligence and its potential to change the world, the technology is still learning, which requires all the data ever created and more. Artists are upset because platforms like ChatGPT are using their art without compensation or even acknowledgment in many cases. Due to systemic racism and misogyny, many of these platforms are being fed corrupt data which means the products it produces are filled with racism and misogyny. Then there is President Musk and his Project 2025 political ambitions, possibly endangering patients under the umbrella of improving their AI. Data privacy is more important than ever, but now, there is more confusion than answers regarding this issue. Once again, all I can say is don’t use Grok or any version of AI created by Twitter, read all news articles regarding data privacy (this includes mainstream media articles), and exercise caution when using AI platforms like ChatGPT because of systemic racism and misogyny, and if it all possible, experiment with more than one AI platform. If you get an icky feeling while using one of these platforms, trust it and walk away.

In my opinion, the hype surrounding AI right now is approaching dangerous and off-putting. Yes, there is tremendous potential to make the quality of life of patients like me better. However, using AI is very much a skill that will need to be taught to the public on a continuous basis. AI will only be as useful as the education and constraints of time on its user. This will require patients and public to invest time and practice into exploring this technology before time becomes the enemy and they need answers sooner than later.

That said there is a danger that the current proponents and supports of AI are not understanding the relationship between the time their jobs and organizations are providing to them so they can experiment with this technology without time being the enemy. There hype is contradictory to the reactive nature of health care in general. The thrill of AI’s potential is counter to the medical necessity of time which plays an important role in any patient journey.

Like I have mentioned a couple times already, healthcare is generally reactive in nature. It’s responding to a particular symptom or pain that was not there yesterday. AI is not a tool one can utilize if you have just fallen off a cliff while rock climbing, just been involved in a car accident, in the middle of a stroke or heart attack, or even being knocked around by a cold or the flu. A patient’s first instinct is to go with what they know because there is a level of comfort there. Utilizing new technology, especially one that requires time and skill with the interaction, will most likely be a second or third thought. A lot more attention needs to be given to the relationship between healthcare and time, I believe, before we can get the most out of AI.

Finally, and at the risk of beating a dead horse, I want to emphasize that AI is a technology that won’t break if you play around with it. Ask it questions, ask it the most silly or stupid question you can think of, then ask another. It might seem silly or a waste of time, but this is probably the best way to learn more about this new technology and its capabilities. This post was designed to show you just that. I wanted to take you with me as I asked it questions, then followed up with more. Although this is a very basic how-to guide, it only took me a couple of hours to create. Now that you know that the key to AI is asking it questions, imagine what you could create if you invested, say, another two hours of your time and specific health data. There are already many stories of moms and dads opening ChatGPT, sharing their kids’ current symptoms and test results, and then identifying a rare disease or condition that 19 specialist doctor visits couldn’t diagnose. The hype surrounding this new technology might be obnoxious right now, but the potential good this tool can provide is life changing.AI will not replace your doctor, but it will help you become better engaged with both your doctor and the healthcare system in general which are both important for one’s health and quality of life. In other words, AI is super cool and you should nerd out on it like the rest of us :)