- Messages
- 34
Has anyone else tried using ChatGPT or other generative AI to analyse their condition and suggest interventions? I would encourage anyone to have gut microbiome tests (and other tests and history) analysed by ChatGPT or similar. It has been very enlightening for me.
TLDR: my CFS was possibly caused by green tea, dark chocolate, and berries. And maybe ChatGPT has shown me how to resolve it.
I have used ChatGPT to analyse the results of a genomic gut microbiome test. I got the testing company to give me a simple spreadsheet with the relative abundance of all bacteria detected that I could then feed straight to ChatGPT.
Some features of my microbiome were:
ChatGPT straightaway identified that the high Akkermansia and Flavonifractor would be the predictable result of my years long daily consumption of green tea (high quality sencha), very dark chocolate (85-90%), and blue/rasp/blackberries and that the other issues would flow from that. I was shocked to think that these 3 things I had been consuming for their antioxidant content and which I thought would help my CFS were actually fuelling gut dysbiosis and probably contributing somewhat (or even largely causing) my CFS. I do not think the sencha/choc/berries were the whole cause by any means though. My CFS seemed to start with a cold/flu-like virus. I had had several periods of feeling mildly CFS-y for months at a time for about 12 years before the full onset. But those high-polyphenol foods maybe set the stage for full-blown CFS and then caused it to get progressively worse.
ChatGPT was then able to suggest treatments, many I had heard of and tried, some I had not: specific prebiotic fibres, specific probiotics, supplements to support the intestinal lining. But it also was able to tell me how to use and combine these treatments effectively. In the past I had used those same treatments in ways which would have frustrated their effectiveness, most notably continuing to consume the sencha, choc and berries would have blunted such interventions. I've only just started on the protocol I developed with ChatGPT, so I can't say whether its advice was correct. It could also be that it is correct about the gut dysbiosis, but that that is not a core part of the mechanism behind my CFS. The protocol is in the attachments for anyone interested in exact details.
The advantage of the GenAI is that it can analyse a huge amount of data (not only my microbiome results but all the research out there on each of the various species and on all the different treatments) and pull out the salient patterns from it. For a human to do this would be extremely time consuming. Since it retains what you have already discussed with it, you can ask it for more and more specific analysis and advice in a way that just couldn't do even if you were lucky enough to have access to a specialist who knew all the research. Of course, I am in no way suggesting that ChatGPT et al. can or should be a substitute for your doctor. But they can be a useful complement, especially in relation to conditions like CFS where the research is so far from mature, and what research there is is not well-known to general practitioners.
I'll be interested to hear about other people's adventures in AI and/or gut microbiome modification!
TLDR: my CFS was possibly caused by green tea, dark chocolate, and berries. And maybe ChatGPT has shown me how to resolve it.
I have used ChatGPT to analyse the results of a genomic gut microbiome test. I got the testing company to give me a simple spreadsheet with the relative abundance of all bacteria detected that I could then feed straight to ChatGPT.
Some features of my microbiome were:
- Extremely high Akkermansia (and, strangely, not the usual muciniphila species);
- Multiple Flavonifractor spp very high;
- Overrepresentation of unknowns like UBA5446, CAG-series;
- Low core butyrate producers: Faecalibacterium prausnitzii, Roseburia spp., Eubacterium hallii all undetected;
- Bifidobacteria absent except B. longum;
- High levels of pathogenic Eggerthella lenta;
- Lactobacillus and E.coli absent;
- Culminating in low diversity.
ChatGPT straightaway identified that the high Akkermansia and Flavonifractor would be the predictable result of my years long daily consumption of green tea (high quality sencha), very dark chocolate (85-90%), and blue/rasp/blackberries and that the other issues would flow from that. I was shocked to think that these 3 things I had been consuming for their antioxidant content and which I thought would help my CFS were actually fuelling gut dysbiosis and probably contributing somewhat (or even largely causing) my CFS. I do not think the sencha/choc/berries were the whole cause by any means though. My CFS seemed to start with a cold/flu-like virus. I had had several periods of feeling mildly CFS-y for months at a time for about 12 years before the full onset. But those high-polyphenol foods maybe set the stage for full-blown CFS and then caused it to get progressively worse.
ChatGPT was then able to suggest treatments, many I had heard of and tried, some I had not: specific prebiotic fibres, specific probiotics, supplements to support the intestinal lining. But it also was able to tell me how to use and combine these treatments effectively. In the past I had used those same treatments in ways which would have frustrated their effectiveness, most notably continuing to consume the sencha, choc and berries would have blunted such interventions. I've only just started on the protocol I developed with ChatGPT, so I can't say whether its advice was correct. It could also be that it is correct about the gut dysbiosis, but that that is not a core part of the mechanism behind my CFS. The protocol is in the attachments for anyone interested in exact details.
The advantage of the GenAI is that it can analyse a huge amount of data (not only my microbiome results but all the research out there on each of the various species and on all the different treatments) and pull out the salient patterns from it. For a human to do this would be extremely time consuming. Since it retains what you have already discussed with it, you can ask it for more and more specific analysis and advice in a way that just couldn't do even if you were lucky enough to have access to a specialist who knew all the research. Of course, I am in no way suggesting that ChatGPT et al. can or should be a substitute for your doctor. But they can be a useful complement, especially in relation to conditions like CFS where the research is so far from mature, and what research there is is not well-known to general practitioners.
I'll be interested to hear about other people's adventures in AI and/or gut microbiome modification!
Last edited:
