Search results

I don’t want to publicly go into details and I am just one person, but given some relatively recent experiences I’ve had nothing has changed for the better in the NHS specialist service in my area.

A quick and perhaps easy one.. Can AfME put up a prominent “Information for Journalists” page with key facts, figures and messages about ME they want to communicate? This seems to work well for other charities when a journalist needs something to copy and paste, er, research, for a news story...

How do we push the ‘do no harm’ message? To me this is key. There’s lots of people trying to do ’things’ but they are often hijacked or derailed or simply misguided. I feel we almost need to start by not doing things. And by recognising it’s ok to say ‘we don’t know’. A lot of harm is done...

Or perhaps how people who think there was a clear trigger think about or describe that? Or just how quickly he can get 100 of us to answer a question! We’re GuinME pigs

I didn’t want to let looking at the results impact my decision so decided to just vote and went for ‘more than two weeks’ as what I would now call ME happened later and ‘more than’ is a very vague definition (which could include more than 2 months!) It is interesting seeing the skew towards no.

Looks like a follow-up to this paper? (now with a disclaimer added as of 10/3/2023) Oxaloacetate Treatment For Mental And Physical Fatigue In Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) and Long-COVID fatigue patients: a non-randomized controlled clinical trial Discussed in this...

Definitely. It’s rubbish that we even have to think about this. Educating the healthcare professionals we interact with is just another challenge we don’t need. I guess one of the challenges in these documents, is that they can’t cover everything, they can’t be one size fits all, so some...

I haven’t been through this all but it’s good to see copying and pasting from the NICE guidelines rather than trying to rewrite things. I didn’t quite read it like that. “Severe” and “at all times” seems to be quite specific. But I can understand the concerns of people who are mild/moderate...

Like others, I find it difficult to be sure. I won’t go into the full history but there was a period of illness either side of ending up in hospital and a few cock ups. After there were ups and downs but I never got better. I’m not sure at what point it became ME. There certainly wasn’t a gap...

Definitely a goal here. However I find it quite funny they are destroying a very profitable business in the process. Ad and web search was there bread and butter but it has been very very much undermined by the proliferation of AI generated content. There may be a future where we are funnelled...

Interesting. Looks like pretty low levels of lots of most of the standard things, sub RDA. But perhaps interesting as it has things like creatine and coenzyme q10 which often seem to pop up being studied independently in more recent studies?

Yes, the idea of what robust data means here is very opaque.

For this prompt I explicitly said the study was flawed and to provide a rebuttal. I get the feeling knowledge of the paper from training data has been picked up here. Given how much on the open web this paper has been discussed it’s probably not great to assume any actual analysis by the LLM...

And here’s the output from my script. Along with a second which was modified to attempt to critique the paper, I didn’t guide it by explicitly telling it the paper was flawed though… —— The PACE trial was a randomized controlled trial designed to assess the effectiveness and safety of four...

I don’t think they’re any good at actual analysis or critical thinking. They tend to echo what is in the paper or common tropes from their training data. We may have some luck if we instruct them to find flaws. But then it’s not really analysis but doing what we ask of them. Does that make...

I think it’s worth experimenting with these tools to see if there’s ways of increasing accessibility to and awareness of long scientific papers for those who cannot or do not want to read them. Whether they’re actually any good at this is of course another question :)

I’ve been experimenting with my own summariser scripts but I haven’t yet hooked them up to use cloud TTS or share these files yet. Here’s example output for these two papers, any feedback to help tailor my prompts very welcome. —— https://www.nature.com/articles/s41467-024-45107-3 This...

I had trying exactly this on my Todo list! From what I’ve heard from others it is extremely good at sounding convincing but can err towards common held arguments or beliefs rather than what is necessarily in the sources. Looks interesting though, especially for high level overviews if you’re...

Interesting, thank you @Jonathan Edwards

Some questions that were going around my head today. What sort of sample size would be needed to replicate this do we think? This has ~1500 people with ME, could a study be done with this (or fewer) people with ME and compared to the healthy controls in the biobank data? Would it be possible...