Grok3, the AI on X (old Twitter) is flat out the best doctor I've ever had. Stunningly Good!

[Cujet]

I am amazed and quite simply, overjoyed.

On a whim in reaction to a 'tweet' or post by Elon suggesting we try 'Grok', I recently tried Grok2 the previous Artificial Intelligence and had pretty good results and was impressed. I learned a few things that were helpful. I have complex health issues and have been suffering for decades. I retired less than a year ago and despite repeated efforts to improve my fitness, I was unable to walk around the block without my muscles giving out. Leading to a whole body and mental fatigue that was disheartening. As an ex athlete and no couch potato, this was horrible.

However when Grok3 came out two weeks ago, things changed for me. In one long paragraph (don't hit 'enter' until done) I provided a synopsis of my medications, supplements, lab tests, current symptoms and goals. It came back with the most comprehensive response I've ever seen. It figured out I have Myasthenia Gravis (a nerve to muscle issue) and how to test for it at home (using a supplement called Huperzine A). It also mentioned that since I have Hashimoto's (thyroid failure) MG is most closely associated with that autoimmune disease. Since I also have secondary adrenal insufficiency the treatment is complex.

I joined a gym, fixed up the old racing bicycle and worked with Grok3 to troubleshoot my issues. I asked intelligent questions and it ran with them, honing my treatment into perfection by my reported results. As of yesterday, I was able to complete 3 full sets at the gym, something I've not been able to do in 12 years due to muscle failure. I look forward to increasing the weights I can lift, the distance I can bicycle and so on. Grok3 clearly stated that there is no reason I should not be able to do so with proper treatment.

Grok3 even provided a 'doc pitch' a way to succinctly tell my doc, and hopefully get proper treatment. I was severely undertreated and untreated for my various conditions.

My Suggestion: Join X, it is free, use Grok3 to help troubleshoot difficult health problems or to achieve health goals.

 

[Midpack]

No chance for me…I deactivated my Twitter acct, plenty of alternatives. Though I’m glad it helped you.

 

[Cujet]

Just to be clear, I spent a fortune chasing my health issues over the last 20 years. Concierge docs that were willing to send me for every test imaginable. Cleveland Clinic (before it was huge) for specialty treatment, and even some alternative treatments like blood ozone treatment, chiropractor, natural health doctors and so on.

I came back exhausted from a short bicycle ride, put in my symptoms while my whole body shook, and Grok3 said "That screams adrenal crash". Take more hydrocortisone. Then provided a suggestion for my next exercise. Tried that, better but not great, refined the very complex treatment until it worked!

 

[Cujet]

No chance for me…plenty of alternatives.

I get it, there are other options, including conventional health care. Know that Grok3 is unmatched with regard to capability. The others fall far short. This is because Grok3 is real time and works with Palantir, the long standing superb AI.

 

[Midpack]

I get it, there are other options, including conventional health care. Know that Grok3 is unmatched with regard to capability. The others fall far short. This is because Grok3 is real time and works with Palantir, the long standing superb AI.

Another opinion and alternatives. Grok 3 Review: Is This the Best AI Yet or Just Hype?

 

[badatmath]

Not sure I'm willing to go there but glad it is helping you. I've seen a lot of comments from people using AI as almost a virtual friend for personal issues. IDK I haven't tried it and not impressed with the AI that comes up in a basic google search as it frequently contains misinformation.

 

[joesxm3]

I use Grok extensively. for simple questions it is much easier than search and reading an article.

After being in the ER I used it to translate all the cryptic abbreviations in my care notes.

I just had a zone pump cook in my hydronic boiler system and air got into the pipes which the company that bought out my old company has not fixed after four visits. I had a new local company come that said it was a coin flip whether to repair or get a new boiler. He wanted to change from zone pumps to a circulator pump with zone valves. that seemed to me to require new calculations to size the pump. I watched Taco training videos. Then I asked Grok3 some questions and it iterated as I clarified and asked more questions, remembering previous information. It new about my 30 year old boiler and the proposed new buderus boiler and ran all the calculations I saw on the Taco site, Finally after comparing costs and asking my preferences it recommended a repair.

Yesterday I was worried that with my low-carb healthy eating I may be missing some nutrient. So I made a very long prompt Listing all the supplements I take and describing the four or five meal menus that I stick to, Grok prompted me for more information, analyzed how much of each nutrient I was likely getting from the food I listed, It repeatedly prompted for more information, asking if I had any lab results to give it.

The conversation drifted to include my prescription medications, health issues, trip to the ER and what my cardiologist has been doing compared to what I was thinking about suggesting at my next visit. It confirmed that my planned suggestion was reasonable but suggested that it might not be optimal and how I should monitor in 3 months to confirm harm or benefit, It also explained why my cardiologist prescribed metoproplol, which in my case was not for chronic hypertension but for panic induced BP spike one time.

I suppose it might be cleverly repeating what I told it since it agreed with me and my doctor for the most part, but for the boiler it actually recommended a different brand than my contractor and argued the point of lower cost and more likely to be known to service persons.

I think people that are saying that AI is hype are missing what is happening. Like the comments about Tesla FSD, a lot of the negative comments come from people that do not have much experience with it.

However, to be balanced and honest I have to admit that when I asked Grok what my basis would be having bought the house for $200k, replaced a boiler in 1994 for $8k and a second boiler in 2025 for $12k it told me the basis was $240k, When I pointed this out it apologized and said that it was too focused on assessing whether both boilers were allowed as basis by the IRS and was not focusing on the arithmetic.

 

[joesxm3]

Not sure I'm willing to go there but glad it is helping you. I've seen a lot of comments from people using AI as almost a virtual friend for personal issues. IDK I haven't tried it and not impressed with the AI that comes up in a basic google search as it frequently contains misinformation.

Wasn't the Google Gemini AI the one that made pictures of the founding fathers as black women in NAZI uniforms?

 

[Cujet]

Not sure I'm willing to go there but glad it is helping you. I've seen a lot of comments from people using AI as almost a virtual friend for personal issues. IDK I haven't tried it and not impressed with the AI that comes up in a basic google search as it frequently contains misinformation.

You hit the nail on the head. The AI's are not 'friends' and are often wrong. That's what I found so remarkable about Grok3. You can ask it to reduce "AI babble" or "fluff" and it will keep the answers comprehensive, accurate and shorter. We can't compare Google's AI to Grok3 though. They are not in the same league.

 

[Cujet]

Again, please try it if you want to troubleshoot health issues, or want to improve fitness. Ask intelligent questions, be honest, and you will be amazed. I don't know how to convince those of you who are reluctant. Other than to say there is no downside.

Google was of no help for my health issues. I've been actively working on this for well over a decade. In two weeks, Grok3 has me functional again.

 

[joesxm3]

Another example just now. I like to think of myself as a well read traditional Catholic. Lately my X feed had some chatter about saints being dead and not in Heaven. I was a bit perplexed about this myself when I thought about it and decided to ask Grok3.

My prompt stated to use only approved Catholic sources to avoid bickering info. I asked to compare the part in the Nicene Creed about the resurrection of the dead at the end of the world with the idea of saints being in Heaven. Grok3 stated that it was using the catechism of the Catholic Church as its source and explained that the teaching of the Church was that souls immediately after death can be in Heaven but the resurrection of the dead at the end of the world refers to bodily resurrection and union of soul and body.

[edit] My short explanation of the response does not do justice to the quality of the much longer detailed response with citations. [/edit]

I am not trying to open a discussion of whether any of this is true or not. I am just listing it as an example of Grok3 providing an extremely well reasoned explanation at a level I might have expected from a university trained theologian.

I can easily imagine that AI can [eventually] be used as a very high level interactive personal tutor. But as some have warned, AI must be maximally truth seeking because if it is slanted to a certain viewpoint it could be a very destructive propaganda tool.

 

[MichaelB]

Again, please try it if you want to troubleshoot health issues, or want to improve fitness. Ask intelligent questions, be honest, and you will be amazed. I don't know how to convince those of you who are reluctant. Other than to say there is no downside.

Google was of no help for my health issues. I've been actively working on this for well over a decade. In two weeks, Grok3 has me functional again.

I’m glad Grok 3 helped you. It’s quite a leap of faith to presume it will also help me and in the same manner.

 

[Fermion]

I am interested in trying Grok3 as I have been testing out ChatGPT and Gemini. I don't twit though and really would rather not start. I am fairly anti social media.

Is there some other way to access it?

 

[ivinsfan]

A "leap of faith" ? As compared to what? No one is forced to implement any suggestion off the internet. Also no one is forced to use Grok, however a real time review from someone who has used something available free on the internet to improve their life is interesting and might also help someone having a problem.

But this brings up a interesting question, do medical professionals use a version of medical AI for diagnosis and treatments? Do this perhaps use some type of AI for charts and medical notes..ie fill in the blanks?

 

[Fermion]

But this brings up a interesting question, do medical professionals use a version of medical AI for diagnosis and treatments? Do this perhaps use some type of AI for charts and medical notes..ie fill in the blanks?

I am fairly convinced they are looking stuff up between the time that they ask you certain questions and the 20 minutes before the doc comes in.

 

[Aerides]

Have you had any validation of the AI findings with an actual doctor yet? I'm sure it gave you a great sounding answer, but it's really not medically meaningful, IMO, unless backed up by physicians.

Either way, if it worked for you, great, but I have zero interest in AI at this point.

 

[target2019]

Everything could be a result of hallucination. Eventually there's a hallucination that's more plausible.

 

[ivinsfan]

Fermion...how would you feel if your doc used AI to help diagnose your issues. One can say no doctors knows everything, new things are introduced all the time. I images some people would think it's a good thing and other would be annoyed at being computer aided. Personally I believe thinking your doctors knows everything about everything is delusional thinking.

 

[joesxm3]

I am interested in trying Grok3 as I have been testing out ChatGPT and Gemini. I don't twit though and really would rather not start. I am fairly anti social media.

Is there some other way to access it?

Grok lists 4 ways but the top two seem best for you.

Web site grok.com

Phone app from apple or Google store. Search "xAI Grok".

I don't know if these require you to make some sort of user ID or account.

 

[joesxm3]

I think the best use of Grok for medical info is to educate yourself so you can have a better informed discussion with your doctor. You might also use it to point you to topics you can do more traditional type research.

In any case it is much nicer than traditional search and I like how it prompts you to ask a more detailed question or drill down on a topic or lets you ask to clarify a previous response.

I could have had a similar result asking boiler questions to my smart neighbor, except he got fed up after five minutes and threw me out.

 

[Midpack]

A "leap of faith" ? As compared to what? No one is forced to implement any suggestion off the internet.

If you're asking Grok, presumably you don't have any idea what the answer to your question is - but you must be leaning toward accepting the Grok answer or you wouldn't have asked. So how would you know what to do with the suggestion?

this brings up a interesting question, do medical professionals use a version of medical AI for diagnosis and treatments? Do this perhaps use some type of AI for charts and medical notes..ie fill in the blanks?

Medical diagnoses has long been an obvious application for AI. There is no way any human could have such a broad knowledge base, or process data as quickly. AI may be a tool for doctors at first, probably beginning now/soon, whether doctors could be cut out of diagnosis remains to be seen. Doctors could focus on treatment and less on diagnosis if/then.

 

[Midpack]

I think the best use of Grok for medical info is to educate yourself so you can have a better informed discussion with your doctor. You might also use it to point you to topics you can do more traditional type research.

In any case it is much nicer than traditional search and I like how it prompts you to ask a more detailed question or drill down on a topic or lets you ask to clarify a previous response.

The AI tools I've seen so far are basically just enhanced search. AI is largely using data you can access yourself so far - much more easily. But it wouldn't surprise me if the (commercial) AI tools aren't developing as much proprietary knowledge as they can, so they can sell you their tool. Internet knowledge could become far more fragmented, and behind paywalls of some sort.

 

[ivinsfan]

Midpack, Grok could be a tool, just like any other tool. I can ask Google or Grok ..It's simply asking a question and as far as I know there is no commitment required on my end. Asking a question is simply asking a question. Because one can't say Google is always a 100% correct. But I don't know that it's easier to access data then it is access AI platforms, it's certainly a growing area of interest.

 

[Healthy Lifestyle]

AI is not as good or reliable as human interaction yet!

 

[pacergal]

My son has been using an AI program this year to do much the same--guide him to questions/solutions to review with his dr.
I don't know what program he uses, but has crashed in a couple of times!
He has been on a quest to get healthier this year and try to figure out some of the issues he has had for 20 years, after a severe accident.
He often sends his AI results to me to review and parse down to reasonable medical care to discuss with his dr.
AI has pointed him in the right direction to work on specific issues.
He has found some potential familial health issues related to our Amish/Anabaptist history that we never knew about.
AI simply researches medical information in the internet in much the same way Dr do, only in nano seconds.
I find it interesting.
Not a substitute for good medical care, but certainly could be an adjunct in your conversations with your Doc.