{{ Several months ago, I got the results back from some routine blood tests, and let’s just say several numbers were a tad too high. My doctor advised “continued diet and exercise” and signed off on the results.
For the past couple of years, though, my numbers had been inching up, and I was frustrated that I couldn’t seem to do much about them. I requested a phone call from my doctor — surely, she had better advice than what she wrote — but she messaged back that if I wanted to discuss my results, I had to set up another appointment.
So, I did what everyone does in this day and age: I turned to artificial intelligence. With low expectations, I typed my lab results into ChatGPT.
As both a physician and a patient, I found the experience startling. Not because ChatGPT dazzled me with its scientific knowledge, but because it behaved the way I wish modern medicine, and its practitioners, still would. }}
When Chat-GPT has a better bedside manner and superior medical knowledge, the medical profession needs to watch out.
I learned from AI that people with Parkinson’s Disease benefit from 100 mg of Vitamin B-3 per day. This is dirt cheap and readily available. I took it and within 3 days I felt better. This was new research which neither my GP nor my neurologist mentioned.
Wendy (I’m a professional researcher so I always do my own research first).
Ok, so just like Japan and Atlantis, I’ve heard of it but never seen it. Where do I get this CHAT-GPT? Thus far I have pumped questions into either Duck-Duck-Go search engine or Google and used their AI function. That’s all I know about how/where to use anything “AI”. And how is it better than the ones I’m already using?
[quote=“rainphakir, post:4, topic:124871”]
Open your play store (Google Play store) from where you get your apps.
Search for ChatGPT… Or Claude or Perplexity.
Download the app.[/quote]
I don’t get apps. I do not go to a place and get an app. But I suppose I could try it here. But, since you did mention the G word… how would this be different / better from the AI function on the Google search engine page?
Just go to ChatGPT.com and start asking questions. Have a conversation with it like it’s a person.
Regular Google search will now lead you to Gemini for any followup questions you may have. It’s pretty good. Don’t treat Gemini as you would a Google search. Again, tell it what you want in conversational form.
Oh I’ve been doing that all along. Proceeding as if I am literally asking an informed person a question instead of slogging thru a “web search” in 1990’s style. It’s great. I love it. And yes, it’s replies are very human.
I just checked, and right there at the Google AI window it mentions Gemini. I get the feeling I haven’t missed anything. Chat-GPT is essentially what I’m getting.
I have noted tho that the Duck-Duck-GO AI is less personable, less informed, and more taciturn than the Google one. Unless it’s just designed to await very specific questions which it can answer with very specific answers. But I’m using AI because I might not know the actual specific question to ask. So, I appreciate the back-and-forth
On the desktop type chatgpt.com, claude.ai or just use Gemini in Google.
Chat is slightly vague responses, but ultra quick.
Claude is detailed and organizing, but slower.
Gemini is organizing, but not well detailed, and quicker.
They are going for different market usages. Claude is outstanding for coders. Gemini wants to be outstanding for shopping. Chat is fudging what it is good for. Chat is satisfying how fast it is.
They all limit how much you get for free. Gemini has higher limits.
Go into settings and control the privacy settings.
But you realise it’s not (an informed person or an authoritative source in and of itself) right?
One of the things I’ve noticed … specifically about Google AI overview as an example, since that seems everso popular on this board as a resource to quote verbatim…is that it somehow manages to bamboozle the questioner into the sort of confidence that overrides any critical thinking skills or desire to fact check in the most rudimentary fashion. Either through confident pronouncements or overt flattery (“Spot on!”)
I have a hard time fathoming how much better AI can perform in the future. Even if this sort of bias confirmation gets remedied movingforward, is information on a topic where one isn’t already somewhat knowledgeable going to be useful … seeing just how much misinformation is sloshing around already and likely to grow exponentially.
It ain’t so much what you don’t know that gets you into trouble … but what you know (courtesy of AI) that just ain’t so. Or, as I say, I’d sooner be underinformed and well aware than misinformed and not.
OMG, @Leap1 …you’re a champ. As I was typing the above response, I was about to say that I am obviously a bit leery of any resource retrieval model that uses me (and a post here on TMF) as its primary resource. That was Google only mind, and, although as time has gone on the replies have used other resources, it’s tended to devolve into examples involving the Law and fictitious precedents etc (forever tarnishing my image of Perry Mason and Horace Rumpole!!)
Now the daughter tends to use ChatGPT to analyse the albatross’s long winded, pettyfogging communications (via his equally high conflict counsel) in order to try to spot a motive/pattern in the ad hominem attacks, filing motions etc., but I’ve never actually asked a question. You provided the easy means so I typed “can domestic abuse evaluators be misled by AI generated false narratives” … it gave me the answers I was looking for! Right down to dressing up the Lump of Foul Deformity’s account into a very plausible story of coercive control by my daughter…complete with examples of all the trendy catchphrases the Plague Sore uses such as “narcissist” and “gaslighting”…the very things that alerted me to something dodgy when I read the whole report. My new BFF.
So, as reluctant as I am to hijack this thread with personal issues, I will anyway. I hope it won’t be necessary for casual readers to need this insight…but but dh and I were once in that position (looking at friends’ anguish over their childrens’ divorces) and thought ourselves unlikely to find ourselves in the same boat. And now we are. Big Time. I wish I knew all this just 2 years ago …
HA! But of course. Only an idiot or someone morbidly attached to a stem degree would get an H/O over this kind of pseudo-wizardry (As in The Wizard of Oz) It’s a big database that can scan really really fast. Faster than I can, and uses familiar speech patterns to present the data instead of old style computer “One-liners.” It’s its “Schtick”. Sound human, friendly, charming, endearing. My GAWD! What an advancement over “List names” Followed by the list of names scrolling
.
In fact I have caught both the Duck-Duck-Go and GoogleGemeini whatever it is, in lies or misrepresented data. Like a salesman who just says stuff. Real, fake, good, bad, just spouts whatever he has to, whatever he’s been told to. With all the confidence of The Burning Bush in the Bible. Presenting it as real or true. I look it up on my own. Not true! I tell it “Hey, that’s not right.” Then it congratulates me on catching the error. “Hey, that’s right! You have a good eye for … (whatever it was)”
I’m glad I didn’t have to rely on gadgets (we actually called the big vacuum tube computer downstairs “The Gadget”) like this back when I was controlling aeroplanes.
But I DO appreciate how fast and comprehensively it can do a websearch especially on a multi part question. It can “think”, for lack of a better word, to check sources I might not have even thought to check. It’s like Wikipedia. Want information? It’s a great place to start. Big bag o’ data in one place. Usually with some references you can search yourself but often dead or outdated links. But don’t take it to the bank.
tiI have GROK, ChatGPT, Claude, Gemini Perplexi and Deep Seek apps in my phone.
Grok can search X quickly. It supposedly will tell more truth than the others. But when you are talking to it it gets very wordy.
ChatGPT works well but it is not a favorite.
Claude is my favorite. It converses well and is pretty accurate but it can be limited in the free version and it can be slow.
Gemini can search the transcripts of Youtube videos. The rest will guess and trying to parse details out of the influencers there. None of the others can do that.
Perplexiti I don’t know what is special about it.
Deep Seek can search Chinese websites directly. The others cannot. It work well actually.
