Where was this advice 18 years ago when I needed it?
OK, so ChatGPT is my new doctor… : Where was this... - Thyroid UK
OK, so ChatGPT is my new doctor…
42 Replies
•
nothing left out.
ChatOn
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 toSneedle
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• in reply to
Can it provide verified sources for meds? That's my central question
• in reply toAlanna012
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• 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 toAlanna012
yes you can always try 🙂
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 toAlanna012
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 🤞
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
TiggerMeAmbassador
What is this advanced voodoo?? Far more useful than a GP visit.... does it prescribe?? 🙏
😆😆
• in reply toTiggerMe
Sadly it doesn’t prescribe… yet!
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 toTiggerMe
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…
Had to google Black Mirror sounds intriguing 🤗
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!
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 😬
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…
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.
Yep... it's the human involvement that rather scuppers everything 😕
Greed generally screws things up
• in reply toTiggerMe
I’ve asked it a few political Qs… it has a very clear leaning. Scary 😟
• in reply totattybogle
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.
Stand out for me "Let us harness its power with wisdom and care" 😳
From the mouth of Ai 🤣
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.
"Essential to consult your healthcare provider" - unfortunately your healthcare provider won't agree with these suggestions so you are stuck!
• in reply toAnthea55
very true…! I’ve become snow blind to warnings like that, they don’t even register with me anymore 😂
helvellaAdministrator
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 tohelvella
Even as such a crude, basic, parroting tool it does MIIIIILES better than the tens of doctors I’ve seen ☹️
• in reply tohelvella
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.
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 😂😂
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.
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.
*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...
My Doctor is Ignoring My Concerns
Good Day,
For two years now, I have been telling my doctor that something is not right with me and...
Is everything ok?
Hi! I am recently diagnosed with Hashi's after years of hypothyroidism. This is my 4th blood test...
Is grapefruit OK?
I'm currently trying to find a solution to my latest hypo related horror-Nocturia (getting up to...
is this ok
I had a total thyroidectomy 11 years ago and I’m on 175 levothyroxine and 60 liothyronine but I...
is this ok?
thyroid uk saidHi
I am assuming that you are not already on treatment.
I can see that your TSH...
Disappointed with my new Primary Care Doctor
Doctor says I am OK then why do I fell so bad
Opinion my doctor
My doctor is refusing to prescribe T3
Doctor says that my thyroid is normal 
