Part 1: Drink Deep, O Luckless
Though the sun ain't yet risen, the morning light still shines through every crack unsealed by our tattered blinds. Just beyond the windows, I can make out the shape of rolling Piedmont hills, their edges soft against the sharpness of our barbed-wire fence, held steady by nothin' but old wooden posts. The wind hisses through 'em, making our cattle huddle up close together for warmth. As I sit on the edge of the bed, my head still pounding from yesterday’s tears, I watch my darling Caroline, still fast asleep.
Last night, I had took the whiskey to bed, and Caroline’s friends sat around the hearth and laughed until the fire burned low. Well, Caroline hadn’t laughed. She just sat there, listening to 'em talk 'bout better days.
I worried 'bout her then. It was the way her eyes looked, lookin’ half lost in her own mind. It weren't too strange for her to be quiet, not really. There was somethin' in the way she moved--a kind of uncertainty. Didn't take long for the others to notice neither.
My feet still hang off our bed as the sun creeps over them hills, stinging my eyes, though I don’t move none. The hallway behind me is still dark. I see Caroline still dreamin' in her slumber, breathing soft in her sleep.
I should have known, Caroline. But I didn’t. Ya didn't say much, just stared off toward the trees like you was waitin' for somethin'--like you knew somethin' the rest of us didn't.
I try to close the blinds tighter, but it’s no use. The cracks are small, but they sure stand out. I wrestle with the cord, but nothin' works. It snaps, drops to the floor and the whole shutter gives way, lettin' the full blaze of the sun in. I reckon ya knew what was comin'. But ya couldn't find the words for it. When ya tried, it came out all confused, nonsensical. Didn't make no sense. Your friends just laughed, thinkin' maybe the booze got to ya. I thought so too.
Then ya collapsed, Caroline. You went stiff as a board when ya hit the floor, shakin' and twistin', your eyes rollin' back in your head. Ain't nobody thought it was the drink then.
Ya had a seizure, and I didn’t know it right then, but I was 'bout to lose ya forever.
Part 2: I Thought We’d Grow Old
Little Molly misses you.
She keeps asking if you’re coming home--back to us, back to normal. I wish I could lie to her. Lord knows I’ve tried. But I can’t shield her from this darkness--not anymore. Not after what the doctor said.
Said that it’s no bigger than a cherry, but it's real mean. A tumor, deep in your brain. And with it already wreckin’ your senses, I suppose there ain’t much hope left.
The hospital's always dim when Molly and I visit. You can smell the decay--the sickness seeping into the cracks of every brick. The mold don’t help much neither, but I guess that’s what you get with a place this old, tucked so deep in the country.
It ain’t the doctor’s fault, knowing he did his best and trusted his judgement. He said if they’d had better tools, they could’ve saved ya.
Kept ramblin’ on about AI. Early diagnosis and the like.
But you weren’t listenin’. Just staring past him, past me.
They recommended moving you to a hospital in the Triangle--said you’d get proper care there.
Said they’ve got AI.
There’s nothin’ they can do to stop it now--it’s spread too far.
But they say the AI might help manage the pain… keep you comfortable in the end.
That’s all they can offer ya now--not hope, just comfort. And even that depends on the doctors knowin’ how to use it right.
Part 3: The Hands that Heal
I don’t trust it.
That’s what I told the big-city doctor when he mentioned AI. The words slipped out before I could think.
He nodded like he’d heard it before.
He told me he understood. Said it’s not perfect--but that it sees what we sometimes miss, especially with patients like Caroline.
She was barely responsive now. Her breathing had changed. Quieter, but strained. She hadn’t spoken in two days. I didn’t know if she even knew I was there anymore.
He said she couldn’t tell us how bad it is. But the AI could read her--subtle changes in her expression, her heart rate, the way her brain reacted. It told them when the pain spiked, even if she couldn’t.
He sounded proud of it.
Like it was some miracle.
Maybe it is.
But I’ve known you longer than any machine. I know you, Caroline.
Still, he looked me in the eye and said they don’t follow it blindly--they use it to listen better.
I wanted to believe him. God, I did.
But I couldn’t help thinkin’--how can some machine know how much you hurt, when I’ve known you half my life and can’t read you no more?
He must’ve seen it on my face. Told me it’s not there to replace them. That AI’s a tool, not a decision.
Said doctors still lead--it just helps them see what they’d otherwise miss.
But it can’t help you. Not now.
I don’t trust it yet.
But I want to.
Little Molly’s here too. She sits beside me, playing with a toy you gave her years before. She drops it. Again. And again. And again.
That’s when the doctor asked if that happened often.
It catches me off guard at first--him suddenly shifting his attention to Molly.
She picks up the toy again, then drops it without looking. Her fingers twitch a little before she lets go.
I nod slowly. Said sometimes. That she says her hand feels sleepy.
He knelt in front of her, gentle as ever, and asked if her arms ever felt tingly--or her legs heavy, like they didn’t want to move.
Molly just shrugged, not really sure how to answer. She held the toy close.
He didn’t press her. Just looked back at me and said there were a few things he’d want to check. Just to be safe. Said sometimes kids her age grow out of little coordination issues--but sometimes, they find patterns worth looking into.
I didn’t say anything. His words just hung there.
He explained that these patterns don’t always show up clearly. Not at first. That’s where the AI helps. Said it could scan her MRIs, her reflexes, and compare them to thousands of other cases instantly. Helps them see what they might’ve missed.
Said they’d run the scans either way--but that the AI was like a second set of eyes. Faster. Less likely to miss the quiet stuff.
Caroline’s still to my side.
Now Molly.
"She’s just a kid,” I whispered.
He told me he knew. Said that’s why it matters--early diagnosis, he meant. Then he stood, giving me space to think.
I looked at Molly. At her toy, now long forgotten on the floor.
AI can’t fix Caroline.
But maybe it can help Molly.
God, I hope so.
Part 4: Lethe Runs Through The Tar Heel State
Sometimes, I feel like the unluckiest man alive, and after looking at your test results, I think I’m beginning to understand why. I’ve made a mess of my life more times than I can count. But until now, there was always a means to fix those messes. This time, there’s nothing left to do but watch you slip away.
Caroline, you lie there, so still now. Your breath is shallow, coming slower with each moment. Your eyes open at times--just a flicker here and there. I reach out, hoping you’ll see me, but they’re already gone, lost somewhere far beyond this room.
I whisper your name sometimes, but I can never tell if you’re there. You never answer.
The house feels colder now, the shadows longer. And I know-- deep down these are your final days, Caroline. Your memory is fading like the sun at dusk. But even if you’re gone from me, I will carry you in my heart. I will carry you into the mornings Molly will live to see--mornings you’ll never get.
Part 5: A Name for the Quiet Things
The house is quiet again, but not like before. After Caroline passed, the silence was heavy and clinged to every surface. Now… It’s different. Still. Focused. Like the breath you take before doing something that matters.
They buried her under the old elm on the hill. The same place she used to read aloud to Molly, voice drifting like wind through the trees. I haven’t been up there since. I’m not quite ready yet.
The clinic ran Little Molly’s scans right away. AI combed through them in under an hour.
It's M.S.
Not a guess. Not a maybe... A name.
There’s things too small for the naked eye. But the machine caught them, patterns too subtle for human perception. They said without AI, we’d have waited a year--maybe longer. By then, the damage might’ve been irreversible.
Thank God they started treatment today. Immunotherapy. A tailored plan--updated in real time with every test result, every change made. The AI’s watching over her like a second pair of eyes we never had. The doctor says Molly’s chances are strong--that the disease is there, yes--but caught early. Managed early. She can still live a full life. Run. Laugh. Dream.
I asked him, all quiet-like, if he really believed that. He looked me dead in the eye and said: “She’s lucky. Most kids like her don’t get seen in time. But this--this is why they use AI.”
It’s strange. Caroline died before the AI could help her. But somehow… It's her who saved Molly. Her death cracked me open. Made me listen. It made me stop fighting the thing I feared, long enough to see what it could do.
It gave Molly a chance.
And tonight, Molly asked if she’d be okay.
For the first time in months, I didn’t have to lie.
I said yes.
And I meant it.
Part 6: Where the Light Comes In
If it weren’t for this house, we’d have been gone long ago--headed to Raleigh maybe, where they’ve got options--Where they’ve got AI, and doctors who know how to use it. Out here, we’ve got a century-old clinic and a prayer.
That’s all there is--a prayer.
The first light slips through the cracks in our blinds--still old and tattered, though it doesn’t bother me now. The house still creaks in the same places, but something about it feels different. Like it’s standin' taller.
She’s outside now, running barefoot through the morning dew. Her laugh carries loudly through the hills, all the way back to the porch where I sit with my coffee. The kind Caroline used to make.
I think about her every morning.
I imagine I always will.
Back at the clinic, they showed me charts, brain scans, predictions. The AI flagged changes in Molly’s condition before she even noticed them. It adjusted her treatment before the symptoms worsened. The doctor said, plain and simple: without it, they’d be guessing in the dark.
Still, it’s not the AI I thank; It’s the man who listened--the one who looked past the code and saw the child in front of him. The one who used the AI like a scalpel--not a compass. Who trusted his own eyes first.
That’s what matters here, in Carolina. Not just having the tools, but knowin' how to use 'em. Knowin' when to follow, and when to lead.
AI doesn’t heal. It didn’t save Molly. The doctor did. He saw what others missed--what I missed. He trusted his training and acted. The AI gave him information--but it was his judgment that made the difference.
The hills are bright now, painted gold by the sun. Caroline would’ve liked that. She always said the world was softer in the morning.
Maybe she was right.
The morning light no longer bothers me now. I let it bore into my eyes just as I did all those months ago, when you slept beside me. I don’t try to fix ‘em now. It’s done. I can’t change what’s been. All I can do is let the light guide me now--let the future take hold.
I know you’re not coming back to me. I’ve made my peace with that. But I will always miss you. You gave our little girl a chance to live a life worth living, and for that, I will always be grateful, even if it means I'll never be whole again--not without you.
Goodbye Caroline.
Work Cited
Bajwa, Junaid, et al. “Artificial Intelligence in Healthcare:
Transforming the Practice of Medicine.” Future Healthcare Journal, vol. 8, no. 2, 2021, pp. 188–194. NCBI, pmc.ncbi.nlm.nih.gov/articles/PMC8285156/
Baxley, Jaymie. “Disparate Issues Shape Rural Health in North Carolina.”
North Carolina Health News, 19 May 2023, www.northcarolinahealthnews.org/2023/05/19/disparate-issues-shape-rural-health-in-nc/. Accessed 27 May 2025.
Ciftci Kavaklioglu, Beyza, et al. “Machine Learning Classification of Multiple Sclerosis in Children Using Optical Coherence Tomography.”
Multiple Sclerosis Journal, vol. 28, no. 14, 9 Aug. 2022, pp. 2253–2262, //pubmed.ncbi.nlm.nih.gov/35946086/. Accessed 27 May 2025.
Concordia University Texas. “States with the Biggest Discrepancies between Urban and Rural Healthcare Access.”
Concordia Texas Accelerated Bachelor’s in Nursing, 5 Mar. 2025, absn.concordia.edu/states-rural-urban-healthcare-disparities/. Accessed 27 May 2025.
De Sario, Gioacchino D., et al. “Using AI to Detect Pain through Facial Expressions: A Review.”
Bioengineering, vol. 10, no. 5, 1 May 2023, p. 548, https://pmc.ncbi.nlm.nih.gov/articles/PMC10215219/#:~:text=Pain%20assessment%20is%20a,this%20area.&text=of%20AI%20in%20clinical,this%20area.&text=current%20state%20of%20the,this%20area.&text=of%20databases%2C%20confounding%20factors%2C,this%20area.
Elevenlabs. “ElevenLabs - Generative AI Text to Speech & Voice Cloning.” Elevenlabs.io, 2024, https://elevenlabs.io/Accessed 29 May 2025.
Golnar Ghane, et al. “Pain Management in Cancer Patients with Artificial Intelligence: Narrative Review.”
Scientifica, vol. 2025, no. 1, 1 Jan. 2025, pmc.ncbi.nlm.nih.gov/articles/PMC12041634/?utm. Accessed 27 May 2025.
IPSpecialist. “The Future of Healthcare: Combining AI Innovation with Human Compassion.”
Medium, 16 Apr. 2025, ip-specialist.medium.com/the-future-of-healthcare-combining-ai-innovation-with-human-compassion-254e6808dd9b. Accessed 27 May 2025.
Key, Alex. “MYCO Medical - NCHA.”
NCHA - Uniting Hospitals, Health Systems and Care Providers for Healthier Communities, 23 July 2021, www.ncha.org/strategic-partners/myco-medical/. Accessed 27 May 2025.
Kumar, Harendra, et al. “Use of Artificial Intelligence for Early Diagnosis of Multiple Sclerosis (MS) and Neuromyelitis Optica (NMO) from Neuroimaging: A Systematic Review and Meta-Analysis (P4-1.003).”
Neurology, vol. 104, no. 7_Supplement_1, 7 Apr. 2025, www.neurology.org/doi/abs/10.1212/WNL.0000000000212346#:~:text=AI-driven%20neuroimaging%20analysis%20has,produced%20the%20most%20accurate%20results. Accessed 27 May 2025.
Rubin, Matan, et al. “Considering the Role of Human Empathy in AI-Driven Therapy.”
JMIR Mental Health, vol. 11, no. 1, 11 June 2024, p. e56529, mental.jmir.org/2024/1/e56529. Accessed 27 May 2025.
Shams, Rayad B., et al. “Community Socioeconomic and Urban–Rural Disparities in Prehospital Notification of Stroke by Emergency Medical Services in North Carolina.” Southern Medical Journal, vol. 116, no. 9, Sept. 2023, pp. 765–771, pmc.ncbi.nlm.nih.gov/articles/PMC10491424/. Accessed 1 Nov. 2024.
Srivastava, Divya, et al. “Exploring the Role of Judgement and Shared Situation Awareness When Working with AI Recommender Systems.”
Cognition Technology & Work, vol. link.springer.com/article/10.1007/s10111-024-00771-9 9, 26 July 2024, Accessed 1 Nov. 2024.
Urban Health Today. “Breakthrough AI Model Detects Multiple Sclerosis Progression Earlier than Clinical Diagnosis | Urban Health Today.”
Urban Health Today, 2025, urbanhealthtoday.com/post/breakthrough-ai-model-detects-multiple-sclerosis-progression-earlier-than-clinical-diagnosis?utm_. Accessed 27 May 2025.
Implicit Evidence Supported
-
AI can detect subtle signs of disease that human doctors might miss.
→ “The AI can read her--subtle changes in her expression, her heart rate, the way her brain reacts…”
Support:
-
AI can detect early-stage Multiple Sclerosis (MS) more effectively and earlier than traditional methods.
→ “Without AI, we’d have waited a year--maybe longer. By then, the damage might’ve been irreversible.”
Support:
-
Machine learning classification of multiple sclerosis in children using optical coherence tomography
-
AI can compare a patient’s data to thousands of other cases instantly to spot patterns.
→ “It could scan her MRIs, her reflexes, compare it to thousands of other cases instantly.”
Support:
-
AI can tailor treatment plans in real time based on test results.
→ “A tailored plan, updated in real time with every test result, every change.”
Support:
-
AI is used to manage end-of-life care, including pain detection and comfort management.
→ “The AI might help manage the pain… keep you comfortable in the end.”
Support:
Claims About the Healthcare System in North Carolina / Rural Areas
-
Rural hospitals have worse facilities and less access to AI-driven tools.
→ “Tucked so deep in the country... the mold don’t help neither...”
Support:
-
Patients in rural areas receive delayed or less accurate diagnoses due to a lack of technology.
→ “They said if they’d had better tools, they could’ve saved you.”
Support:
-
AI tools are more common and effectively used in advanced hospitals, like those in Raleigh or the Triangle.
→ “If it weren’t for this house, we’d have been gone long ago--headed to Raleigh maybe, where they’ve got AI.”
Support:
Claims About Human vs. Machine Judgment
-
AI should be used as a tool to assist doctors, not replace them.
→ “AI’s a tool, not a decision... It’s the man who listened.”
Support:
-
Human compassion and judgment are still essential in AI-assisted care.
→ “The one who used the AI like a scalpel--not a compass.”
Support: