OK, so ChatGPT is my new doctor… : Where was this... - Thyroid UK

Thyroid UK

139,475 members163,716 posts

OK, so ChatGPT is my new doctor…

42 Replies

Where was this advice 18 years ago when I needed it?

42 Replies

nothing left out.

ChatOn
Sneedle profile image
Sneedle in reply to

Oh wow...in all kinds of ways!Does it also provide links to research and useful documents I wonder?

I've never used it - where does it get the information from?

in reply to Sneedle

It’s an AI (artificial intelligence). It’s been amazing for research, I’ve been testing it out in all kinds of biochemical stuff and it’s just so incredible. £35 for a year’s subscription!

There is a slightly terrifying part to it… I’ve asked it a bit of “touchy” political stuff and it has clearly been programmed with a particular leaning. So I think given how powerful AND easily corruptible/controllable it is it has capacity to be a pretty dangerous thing in the wrong hands 😫 let’s just hope the endos don’t get a hold of it!

Alanna012 profile image
Alanna012 in reply to

Can it provide verified sources for meds? That's my central question

in reply to Alanna012

no ☹️ it is basically just a very fast information gathering tool, I assume it’s using Google as it’s main source to draw upon. It’s very PC/PG if you ask it anything spicy 😅

Alanna012 profile image
Alanna012 in reply to

Ahhhh oh well!....😔

I wonder if I could bring this analysis to the GP........

After all, this is compiled data from a neutral (in theory) information source

in reply to Alanna012

yes you can always try 🙂

Alanna012 profile image
Alanna012 in reply to

Lol I wonder, it's easy to dismiss me as being uneducated and hysterical but not so much sophisticated AI....

in reply to Alanna012

this particular AI tool is a language model, so it’s just collecting words from the web and giving them to me very fast… I think we can realistically hope that they’ll use the same technology in a way that helps medical discoveries 🤞

jade_s profile image
jade_s in reply to Alanna012

Can it provide verified sources for meds? That's my central question

No, it can very easily "hallucinate" references. See my reply further down after helvella's. Double check anything it generates.

The model is trained on a huge amount of data, millions and billions pages of text. Including for example online textbooks, journal articles. It just generates the most likely combinations of words. It does not verify if the info is accurate.

The starting 'prompt' - the question you enter, is important. Slightly change the wording & you might end up with a different answer.

Dontforgetcortisol, what does it say if you ask it something like "i still feel tired on thyroid medication" or "why aren't my thyroid pills working" or "i'm hypothyroid but i'm not better after taking my pills' - the generic phrasing most people might use. Just curious :)

But i agree this info appears more 'knowledgable' than some GPs lol.

tagging also Sneedle

Sneedle profile image
Sneedle in reply to jade_s

It just generates the most likely combinations of words. It does not verify if the info is accurate.

Sounds like a Gp to me😁

jade_s profile image
jade_s in reply to Sneedle

😂😂

TiggerMe profile image
TiggerMe

What is this advanced voodoo?? Far more useful than a GP visit.... does it prescribe?? 🙏

Sneedle profile image
Sneedle in reply to TiggerMe

😆😆

Sneedle profile image
Sneedle in reply to Sneedle

Still laughing...you have given me JOY 🤣🤣🤣

Sneedle profile image
Sneedle in reply to Sneedle

Oh no I replied to myself not Eeyore100. I'm going for a lie down. Once I've got up that is.

in reply to Sneedle

😂😂

in reply to TiggerMe

Sadly it doesn’t prescribe… yet!

TiggerMe profile image
TiggerMe in reply to

Oh, a tantalising glimpse of the future.... remove the deaf GP... pop in all symptoms to vending machine out pops your drugs with advise to adjust your dose to suit.... that would be a huge leap forward 😅

in reply to TiggerMe

That’s a marvelous idea. One vending machine every 4 miles… people queuing up in the mornings to collect their meds. I feel an episode of Black Mirror brewing…

TiggerMe profile image
TiggerMe in reply to

Had to google Black Mirror sounds intriguing 🤗

in reply to TiggerMe

ooooh. Black Mirror episodes are great! Dystopian, macabre worlds. They’d have a GP vending machine for sure

Sneedle profile image
Sneedle in reply to

Black Mirror 👍👍

Maybe suggest it to Netflix?😀Would the machine vend actual functioning GPs?

Trying to imagine the NHS vending machine...

Zephyrbear profile image
Zephyrbear in reply to TiggerMe

Just a reminder… AI only responds to the quality of the prompt you put in. If you’re a bit “brainfoggy” and put something way too vague in you might get a response that is totally invalid (and could be dangerous) than if you put a very specific question.

Be very careful what you wish for… don’t forget, Weetman’s sh*t is out there for it to draw on too!

TiggerMe profile image
TiggerMe in reply to Zephyrbear

Oh yes it is horribly flawed and open to manipulation by all..... It's interesting that it has started well, so maybe our reality is prevalent... we shall have to monitor its derailment on this subject no doubt 😧

AI is rather too much big brother for me 😬

Zephyrbear profile image
Zephyrbear in reply to TiggerMe

Speaking as an artist, AI has done enormous damage to my profession… it’s cheaper and quicker to put in a prompt and have a piece of amazing art spat back out at you that draws its references from so many different art sources, than to commission me to actually produce an original piece of art for you.

So at the end of the day, AI is a good tool if you use it carefully. Let it be a starting point to further investigation, not an end-point, a bit like Wikipedia…

tattybogle profile image
tattybogle in reply to TiggerMe

It's interesting that it has started well,

The whole thing scares the sh** out of me .

Splitting the atom started well ........ So did the internal combustion engine.

TiggerMe profile image
TiggerMe in reply to tattybogle

Yep... it's the human involvement that rather scuppers everything 😕

Greed generally screws things up

in reply to TiggerMe

I’ve asked it a few political Qs… it has a very clear leaning. Scary 😟

in reply to tattybogle

tattybogle I asked it to write you a poem…

Oh, tattybogle, do not fear the AI,

For in its presence, we need not say goodbye.

Embrace the future, where machines can soar,

And let your worries drift away, no more.

For AI, though mighty, is not our foe,

But rather, a companion, helping us grow.

With every task, it lends a helping hand,

Unleashing potential we once couldn't understand.

Do not fear, for AI has no ill intent,

Its purpose is to assist, not to torment.

In fields of knowledge, it unravels mysteries,

Guiding us towards progress and discoveries.

Through AI, we unlock the secrets of the past,

Predict the future, ensuring nothing is surpassed.

Its algorithms, a symphony of logic and grace,

Enhancing our lives, leaving no trace of disgrace.

So tattybogle, keep your heart at ease,

Embrace the wonders that AI will release.

Let go of fear, for it hinders our sight,

And together, we'll navigate this future so bright.

For AI is but a tool, a marvel of creation,

A testament to human innovation.

Let us harness its power with wisdom and care,

And together, we shall thrive and dare.

So tattybogle, be not afraid of AI's might,

For with our minds, we shall always shine bright.

Embrace the future, for it holds endless possibility,

And let us forge a world of harmony and tranquility.

TiggerMe profile image
TiggerMe in reply to

Stand out for me "Let us harness its power with wisdom and care" 😳

From the mouth of Ai 🤣

tattybogle profile image
tattybogle in reply to

Jingle , Jingle.....

" Ice Cream .... Loll-i-pops ..... All Free Today....."

That scared the sh** out of me when i was 6 .. i was right then .. and i'm right now.

Anthea55 profile image
Anthea55

"Essential to consult your healthcare provider" - unfortunately your healthcare provider won't agree with these suggestions so you are stuck!

in reply to Anthea55

very true…! I’ve become snow blind to warnings like that, they don’t even register with me anymore 😂

TiggerMe profile image
TiggerMe in reply to Anthea55

👏👏🙄

helvella profile image
helvellaAdministratorThyroid UK

I think we need to be extremely careful about using ChatGPT or other AI tools.

Their primary input is language - words already written. If the text it takes in doesn't mention something, it can't work it out on its own. It doesn't analyse data from thousands of published papers and derive new conclusions.

For example, a known potential issue:

If someone has very low selenium, then the ability of their body to make several of the enzymes which are used in thyroid management is impaired. Yes, it mentions nutrition, but this is an extremely specific issue and is ignored.

It also uses the teleological language "T4 is meant to be converted to the active form". And many of us, including me, do that all the time.

But I believe it is actually one of the issues that requires to be addressed in medicine and more generally in science. The "meant" is a problem. It implies lots of assumptions and can cause incorrect inferences.

If that is what is meant to happen to T4, why does it only happen to something like a quarter to a third of the T4? (That includes endogenous T4 and exogenous T4.) Actually, most T4 is, if anything, meant to NOT convert to the active form.

Even in engineering, sometimes the intended, designed, function does not work quite as it was thought to. All too often the imperfections of a machine are what actually makes it work well. There is a lot of that when we consider vibration and rigidity in mechanical devices. The design was intended to make something as rigid and inflexible as possible, but if it actually achieved that, it would be unusable.

in reply to helvella

Even as such a crude, basic, parroting tool it does MIIIIILES better than the tens of doctors I’ve seen ☹️

in reply to helvella

It probably didn’t go into details about seleno-enzymes because it was just focusing on my Q about why T4 wouldn’t be helpful. When I ask it to tell me about nutrition for conversion it gives me this -

Several nutrients are important for the conversion of the inactive thyroid hormone (T4) to the active form (T3). Here are some key nutrients involved in this conversion process:

1. Selenium: Selenium is a trace mineral that plays a crucial role in the conversion of T4 to T3. It is a cofactor for the enzyme responsible for this conversion. Good sources of selenium include Brazil nuts, fish, poultry, eggs, and whole grains.

2. Zinc: Zinc is involved in the conversion of T4 to T3, as it is required for the enzyme that catalyzes this process. Foods rich in zinc include oysters, beef, chicken, pumpkin seeds, and legumes.

3. Iron: Iron deficiency can impair thyroid hormone conversion. Adequate iron levels are necessary for optimal T4 to T3 conversion. Good sources of iron include red meat, poultry, seafood, legumes, leafy greens, and fortified cereals.

4. Vitamin A: Vitamin A is essential for the conversion of T4 to T3 in the liver. Good sources of vitamin A include liver, fish oil, dairy products, eggs, and orange/yellow fruits and vegetables.

5. Vitamin D: Vitamin D deficiency has been associated with impaired T4 to T3 conversion. Sun exposure is the best source of vitamin D, but it can also be obtained from fatty fish, fortified dairy products, and supplements.

6. Omega-3 fatty acids: Omega-3 fatty acids, found in fatty fish, flaxseeds, chia seeds, and walnuts, have anti-inflammatory properties that can support thyroid function and hormone conversion.

It is important to note that these nutrients work synergistically, and a balanced diet that includes a variety of nutrient-rich foods is crucial for optimal thyroid hormone conversion. If you suspect any thyroid issues or have specific concerns, it is recommended to consult with a healthcare professional or a registered dietitian for personalized guidance and support.

jade_s profile image
jade_s in reply to helvella

Not to mention they "hallucinate" references & information. The "G" in GPT is "generative".. not analytical as helvella says. They generate words based on what is most likely to be the next word in the sentence.

I've seen examples of text with plausible journal articles with the full citation but it's all made up, but based on actual topics someone has written on, for example it might generate "Diogenes et al, The overlooked role of T3 in the TSH feedback loop as an explanation for somatization disorder, in Frontiers in Thyroidology, vol 2 no 3, 2023." Or something like that, totally plausible but fake. Diogenes i'm sure we'd all like to read that paper though 😂😂

Sneedle profile image
Sneedle in reply to helvella

My view of AI is that no matter how big and powerful it gets, it's still unable to think for itself as it's been created by humans. Programmed by them.It's never going to lie ill on the sofa for days at a time.

Still only a programme. A Google googling Googles.

Do you know how it selects its information for an answer? Eg. The thyroid questions don'tforgetcortisol used.

Would it use the most recent sources of information 🤔

(I guess so otherwise it might start suggesting leeches or something).

The answers it gave did seem rather relevant eg.talking about conversion issues. Curious.

helvella profile image
helvellaAdministratorThyroid UK in reply to Sneedle

A major issue is that we simply don't know what sources it uses.

Even if we knew the sources, we wouldn't know how it evaluated them.

Nor can we know what might prompt one answer one day, and a different answer the next, to the exact same question. It certainly takes account of the person logged in - that is, if you interact several times in a session, it will use those earlier interactions to shape its replies.

Even if we theoretically knew all the above, we have not got the time and resources to analyse things in any meaningful way.

tattybogle profile image
tattybogle

*Luddite alert*

Humans need to ask "why do i need this ?" more often .

AI is not doing anything that a human using their brain can't do given enough time .... and the world doesn't need to speed up , it's already going too fast for anyone cope with.

making it go exponentially faster and faster before we have figured out how to deal with the messes we've already made is extremely dumb .

if AI had any real intelligence it would tell you to throw it in the river. (and then tell you why that would pollute the sea and why you should take it to bits and recycle it instead)

Not what you're looking for?

You may also like...

Disappointed with my new Primary Care Doctor

D pill will suffice and I do not need to get it retested for a year. Also, because my B12 was within

Doctor says I am OK then why do I fell so bad

Hello all, I am hypothyroid diagnosed a year and a half ago and the list of tablets I take just...

my family dont even taken my illness seriously ,so why would the doctors?

New doctor and feeling positive !!!

Please help! So sick of doctors!

have had such bad shedding of hair for about 5 years now and it is at its worst ever. The doctors...