Skip to Main Content

Advertisement

Skip Nav Destination

From a Chatbot

ASA Monitor Today

It’s increasingly evident that chatbot technology is going to have a lot to say about how we conduct our business and personal lives going forward. This section features updates and observations about these rapidly evolving AI tools.


2024 – March 26

2023 – December 20  |  July  |  May  |  April  |  March


explain this to me like i'm a tenth grader...

March 26

If you’re not using it already, I highly encourage you to get acquainted with Google’s generative AI chatbot Gemini. I won’t bother you with a lot of editorializing, except to say that chatbot technology continues to amaze me. In fact, many of the answers I’ve received from both simple and complex questions make me question the very definition of “intelligence.” 

I recently asked Gemini to “describe the mechanism of action of sevoflurane to a teenager.” The result was a reasonably accurate description, and it was phenomenally clever. Also, it took less than 5 seconds to generate: 

Absolutely! Here's how I'd describe the mechanism of action of sevoflurane to a teenager:
Think of your brain as a giant party:
All the different parts of your brain are guests at the party, and they chat with each other constantly to keep things running smoothly. This chatter is basically electrical signals traveling along pathways.
Parties need the right atmosphere. Sometimes you want them calm, other times more lively. Your brain is the same – it needs the right balance for you to feel awake, focused, or sleepy.
Sevoflurane is the ultimate party crasher:
Sevoflurane is an anesthetic, meaning it's a medicine used to put people into a temporary sleep-like state during surgery.
It basically slows down the communication between different parts of the brain. The chatter turns into whispers, which makes your brain switch into a slower mode.
Even though it sounds a bit scary, this is actually a good thing during surgery, as it means you don't feel any pain and you don't remember the procedure afterward.
How exactly does sevoflurane do it?
Your brain cells (the guests at the party) "talk" to each other through tiny doorways. Sevoflurane likes to hang out at these doorways.
It changes how these doorways work – some become harder to open, and others stay open too long. These changes disrupt the usual chatter, making it harder for brain signals to travel.
The result:
You become less aware of your surroundings. Thinking and reacting gets slower.
You feel less pain. The pain signals struggle to get to where they need to go.
You relax and eventually fall into a deep sleep.
Important Extra Bits:
Doctors carefully control how much sevoflurane you get, so it's safe and you wake up properly.
Once the surgery is over, it stops being given to you, and it leaves your body – that means the brain party can get back to normal!

Let me know if you'd like more details!

Then, for fun, I asked Gemini to render the above request in iambic pentameter:

Your brain, a feast of thoughts both loud and bright, Where signals dance in ever-changing light. Each cell a guest, with whispers all their own, They shape your world, on paths and roadways grown.
Sevoflurane, a sleep no sleep has known, It slows the hum, where thoughts were freely sown. The chatter dims, a hush across the land, Your brain now shifts, subdued by unseen hand.

It clings to doorways where the signals pass, And alters how the messages amass. Some doors stay shut, some linger far too wide, The flow of thought now takes a softer tide.
Less pain you feel, awareness starts to fade, Your senses dull, as slumber's plans are laid. A deeper sleep descends, a gentle night, Controlled by those who guide you towards the light.
And when the task is done, the drug retreats, Your brain awakens, normalcy completes. The party stirs, the guests begin to rise, And so once more, your conscious self takes skies.


Finally, I couldn’t resist asking Gemini to revise its sevoflurane description in the style of David Attenborough: 

Imagine, if you will, the extraordinary landscape of the human brain. A vibrant tapestry of cells, each one a tiny marvel, flickering with the electric pulse of thought. These pathways form a bustling metropolis, humming with the ceaseless exchange of information – the very essence of consciousness.

But now, a silent change descends. Sevoflurane, a potent anesthetic, seeps into this intricate realm. Like a phantom mist, it blankets the bustling junctions where thoughts are traded. Its presence lingers, subtly altering the intricate mechanics of communication.

Some neural doorways, once flung wide open, now find themselves hindered, the flow of signals subtly slowed. Others become oddly fixated, held ajar for longer than intended. The grand symphony of the brain falls into a softer rhythm.

The effect is profound. Awareness dims, like a distant shoreline receding from view. The sharp sting of pain fades, its urgent messages blunted, unable to reach their destination. A deep stillness descends, a controlled slumber that echoes the tranquility of the natural world.

Yet this intervention is precise, measured. Skilled hands guide its ebb and flow, ensuring a safe journey into oblivion and a gentle return. As the sevoflurane dissipates, so too does its influence.  The vibrant metropolis of the mind stirs back to life, circuits reconnecting, the symphony of consciousness resuming once more.

↑ Back to top

Off-label use for ChatGPT?

 

February 15

A newer large-language model (LLM) significantly achieved better scores than human test takers when presented with questions similar to those on neurology board examinations, according to an article in Neurology Today.

The study was not done to see if artificial intelligence (AI) could replace doctors—or take their tests for them—but rather to determine whether LLMs might be useful in real-world clinical settings as well as in medical education.

The answer that emerged from the study seems to be a cautious yes, though more research is needed to fine-tune the application of LLMs specifically for neurology, according to the study investigators, based at University Hospital Heidelberg in Germany.

The study, which evaluated two versions of LLMs—ChatGPT-3.5 and ChatGPT-4—found that the newer version achieved higher scores on board-style exams than the average human scores. Both versions of LLM performed better on lower-order questions—multiple-choice questions, for example, on which drug to prescribe for a given condition—than higher-order ones, which required analysis and synthesis of information.

↑ Back to top

Patients wary of doctors’ use of ai

December 20

The vast majority of American patients are wary of how their doctor may use generative AI to help treat them, according to a new Wolters Kluwer Health survey.

The technology is still in limited use in physician offices — mostly to help with administrative tasks — but one day may help doctors make diagnoses or develop care plans.

The information services company’s new survey indicates that transparency about where the information churned out by generative AI comes from can help reduce patients’ worries.

Roughly four in five say they have concerns about their provider using generative AI to make diagnoses or set treatment, with the vast majority saying that’s because they don’t know where the information the tech uses comes from or why it should be trusted.

Roughly half say they’re concerned generative AI might produce false information, while eight in 10 said they worry the information may come from unfiltered internet searches. While about 80% said they were concerned about their providers using the technology, that dropped to 63% if they knew their doctor was using a tool that came from an established health care source that was constantly being updated. 86% said they would need to know medical professionals were involved in the process of creating source content to feel comfortable with their doctor using generative AI.

“As the health care community begins implementing GenAI applications, they must first understand and address the concerns Americans have about it being used in their care,” said Greg Samios, CEO of clinical effectiveness at Wolters Kluwer Health.

↑ Back to top


July 2023

AI with ourselves as guinea pigs

July 28Brown haired person avatar, LARA

Our ASA Monitor publisher, Wolters Kluwer, has created a chatbot based on their journal and publication content. The software allows users to enter a query and will generate a response from all Lippincott publications within the past five years. The chatbot, LARA (Lippincott Advanced Research Assistant) is currently in beta mode. You can access it at lww.com/pages/ask-Lara.aspx. Wolters Kluwer welcomes your participation in testing it out during the beta release and in providing feedback to help them improve their AI assistant.

↑ Back to top


May 2023

Love It or Hate It, ChatGPT is Here to Stay

May 4

Abstract 1
Abstract 2

ChatGPT may not be taking our jobs, but it is definitely going to be a disruptor. I’ve enjoyed getting to know the technology and have been pretty impressed by its capabilities.

Many colleagues and journals are banning the use of ChatGPT. I think fighting technology is invariably a losing proposition. When a new technology is introduced, one has to use it, understand it, and figure out how to adopt it.

In that vein, ASA Monitor has adapted a policy on ChatGPT which can be found here. This may evolve as time goes on, but I believe that—like Wikipedia in the early days—anyone who fights ChatGPT will be unable to compete with those who understand it.

I’m curious to hear more of my colleagues’ thoughts on ChatGPT and invite you to participate in our Twitter poll. We’ve published a side-by-side comparison of one of my all-time favorite abstracts with one generated by ChatGPT. Please vote on which you think is the original and which was generated by the bot.

P.S. I’d also love to hear your thoughts on the social message itself—it was written by ChatGPT!

↑ Back to top


April 2023

ChatGPT-4’s thoughts on desflurane

April 13

Doctor in scrubs and surgical maskWhen asked whether desflurane should be used, Chat GPT-4 did a reasonable job of summarizing the benefits and concerns about desflurane. However, Chat GPT-4 equivocated about whether the benefits (marginally faster emergence) outweighed the fairly striking environmental effects of desflurane. It seems that environmental risk is presently best left to the humans.

Desflurane is a volatile anesthetic agent used in general anesthesia. It is known for its low blood-gas partition coefficient, which allows for rapid induction and emergence from anesthesia. However, concerns have been raised about its environmental impact and potential adverse effects on patients. The decision to continue using desflurane depends on several factors, and it's important to weigh its advantages and disadvantages.

Advantages of desflurane include:

Rapid induction and emergence: Due to its low blood-gas partition coefficient, desflurane enables quick anesthesia induction and emergence, which can be particularly beneficial in outpatient or ambulatory surgery settings where rapid recovery is desired.

Precise control: Desflurane allows for easy and precise control of anesthetic depth, which can lead to improved patient management during surgery.

Minimal metabolism: Desflurane undergoes minimal metabolism in the body, reducing the risk of toxic metabolic byproducts.

Disadvantages of desflurane include:

Environmental impact: Desflurane has a high global warming potential, and its release into the atmosphere contributes to greenhouse gas emissions. This has led to concerns about its environmental impact and the need for more sustainable alternatives in anesthesia practice.

Respiratory irritation: Desflurane can cause airway irritation and coughing, particularly during induction of anesthesia. This can be problematic in patients with reactive airways or pre-existing respiratory conditions.

Cost: Desflurane is generally more expensive than other volatile anesthetic agents like sevoflurane or isoflurane, which may impact its use in cost-sensitive health care settings.

Ultimately, the decision to continue using desflurane should be based on a careful assessment of its benefits and drawbacks, as well as the specific needs and risk factors of individual patients. Anesthesiologists should consider the environmental impact of their choices and strive to reduce their carbon footprint where possible. Alternative anesthetic agents, such as sevoflurane or isoflurane, may be more suitable for certain patients or in situations where minimizing environmental impact is a priority. It's essential to stay informed about current research and guidelines to make the best decisions for patient care and environmental responsibility.
- ChatGPT-4

↑ Back to top

Our patients are already using it – so be prepared

April 13

Hands working on a laptopEven this paleo-anesthesiologist tried out Chat GPT. I did it to see what it would say about risk of pediatric anesthesia in a healthy child, as well as the risks of dying compared to driving. The responses were reasonable and accurate. (This is the prompt I used for the latter: “For a healthy child, which has the higher chance of death – general anesthesia or a car accident”). Even if we don’t use it, Chat GPT will be another online tool parents and patients will use. So, we need to be familiar with it. The major problem with Chat GPT is the response did not include any references or where the information was retrieved, unlike a Google search. I doubt that the layperson will ask ChatGPT to “find references” on a question and then look at the initial source. Each time I tried it, the references weren’t accurate.

↑ Back to top


March 2023

An AI ode to ASA

March 30

HAL 9000 from A Space OdesseyYou would have to be living in a cave to have not heard of ChatGPT, the large language model recently released by OpenAI. ChatGPT has been heralded as a breakthrough in artificial intelligence. I will be writing regularly about ChatGPT for ASA Monitor Today. While ChatGPT may not be truly intelligent, it does a remarkable job of impersonating intelligence.

For this inaugural note, I asked ChatGPT to write an ode to the American Society of Anesthesiologists. In its own words…

I. Prelude

O noble guild of healers, wise and true,
Whose artful hands do guide the patient through,
The maze of fear and pain, a journey fraught,
We sing an ode to you, the humble thought.

II. Slumber's Keepers

American Society of Anesthesiologists,
Sentinels of sleep, in hallowed halls,
With skillful touch and whispered lullabies,
You guide us safely through the darkest skies.

By your hands, the scalpel finds its mark,
As you navigate the realms of dark,
In trust we lay our lives, as on an altar,
Our fears assuaged, your calming voice a salve.

III. Monitors of Mortal Life

Unseen by many, yet your work profound,
You cradle life in stillness and in sound,
Monitoring breath and heartbeat's tender dance,
In that fragile space where life and dreams advance.

IV. Alchemists of Sleep

Masters of molecules and chemistry,
You conjure potions, mixtures to set us free,
From pain and fear, you guide the surgeon's blade,
A gentle reprieve in this world you've made.

V. Guardians of the Dreaming

In slumber's arms, your watchful gaze protects,
Each patient's dreams, your noble charge, reflects,
A sacred trust, from which you do not waver,
With steadfast hearts, you grant us this great favor.

VI. The Ascent

And when the time is ripe, you lead us back,
From dreams and shadows, up the winding track,
To sunlight's warmth and loved ones waiting there,
Your careful hands restore us from despair.

VII. Homage to the Healers

American Society of Anesthesiologists,
To you, we raise our voices high in praise,
For in your skillful hands, our lives we trust,
With gratitude, we sing this ode to you, a must.

- ChatGPT-4

↑ Back to top

Close Modal

or Create an Account

Close Modal
Close Modal