Will AI programs/models significantly aid in producing viable treatments for ME/CFS?

Will AI programs significantly aid in producing viable treatments for ME/CFS?

  • Yes

    Votes: 6 37.5%
  • No

    Votes: 7 43.8%
  • No idea

    Votes: 3 18.8%

  • Total voters
    16
Also discussed in this thread.

 
I found this difficult to answer because of the significantly qualifier and different people’s understanding…

But in short, they already are, just not in the way public discussion often goes.

DecodeMe used AI/ML. Other papers on here use AI/ML and AI/ML is being used in labs for things like protein design (see the work of the Baker Lab or this podcast, or consider that the output of AlphaFold and other tools has fed into databases routinely used in biology). So in some way these tools will undoubtedly have an impact.

But no a straight commercial chatbot LLM will not. Asking them to answer questions like this is not a good use case IMHO.
 
Last edited:
I found this difficult to ask because of the significantly qualifier and different people’s understanding…

But in short, they already are, just not in the way public discussion often goes.

DecodeMe used AI/ML. Other papers on here use AI/ML and AI/ML is being used in labs for things like protein design (see the work of the Baker Lab or this podcast, or consider that the outpit of AlphaFold and other tools has fed into databases routinely used). So in some way these tools will undoubtedly have an impact.

But no a straight commercial chatbot LLM will not. Asking them to answer questions like this is not a good use case IMHO.
Snap!
 
I just see AI and ML tools as computer systems, I find it hard to think about them as things on their own. They're sophisticated systems, but evolution in computing is so rapid that what we have now will look slow and unreasonably resource-heavy within about 10 years.


ETA:
At school we had to open a line on an old dial phone to the polytechnic mainframe, and if you wanted it to do anything at all you had to write the code—software wasn't much of a thing in the 70s. I remember having to spend weeks working out the code for a set of traffic lights operating a four-way system, after which I decided I'd rather set my toenails on fire than ever go near a computer again.
 
In the event that we have proof of the biological mechanisms of ME/CFS and it's not immediately obvious what drugs could target it, it could end up being useful. And if I was to find some genes or pathways with potential relevance in ME/CFS from an actual experiment, it might save me some time getting some general context for where else those terms have come up in research.

But it will not derive any better "understanding" of ME/CFS, or what drugs could potentially work for it, than what a below-average biology student could haphazardly guess after 3 minutes of googling.
 
My problem in responding to the poll is ‘what time scale are we talking about’?

Ultimately AI will give us important answers, but I feel that is at least years if not decades in the future.
 
Back
Top Bottom