axiomaticdoubts

There've been multiple studies that gave people potassium supplements and measured their body weight. (Primarily because potassium helps lower blood pressure). They haven't found that potassium causes weight loss. See my meta-analysis here: https://docs.google.com/document/d/1sR1T2Kb1X1fCLYeEE-U3Rwei...

I think it's worth pointing out that they've also said multiple false things in their blog posts, and have refused to address or fix them despite repeated requests over a period of more than one year. I think people should be careful before believing in things they read in SMTM's blog. See e.g. https://manifold.markets/Natalia/how-many-of-these-falsemisl...

I don't think mold_time is a good example of citizen science done right. They have said multiple false things and refused to address or correct their statements when other people pointed them out. See https://www.lesswrong.com/posts/NRrbJJWnaSorrqvtZ/on-not-get... and https://www.lesswrong.com/posts/7iAABhWpcGeP5e6SB/it-s-proba...

It's been more than a year since I alerted them of the multiple falsehoods in A Chemical Hunger, repeatedly, and they haven't done anything about them.

The main issue with this riff trial is that it doesn't test the most likely reason the potato diet works: that it's a very restrictive diet. Testing several different hypotheses barely helps at all if you don't test the overwhelmingly most likely one.

The attrition rate was very high. The claim that it's easy was based on a few anecdotes, which aren't representative of the trial participants.

"A Chemical Hunger" has multiple inaccuracies that have never been addressed or corrected. For example, they made up the "fact" that wild animals have been getting fatter out of thin air. See https://www.lesswrong.com/posts/NRrbJJWnaSorrqvtZ/on-not-get...

The critical temperature is not 127 C. They did not actually measure the critical temperature. They just said it's at least 400 K (~127 C), probably because that was the maximum temperature the equipment they used could measure.

People publish blog posts for several reasons. Many see it as a fun, social activity.

It's not just bloggers. There are computer scientists with publications in top CS conferences that post on LW. For example, Alexander Turner, Dan Hendrycks, Paul Christiano, and Jacob Steinhardt all post there.

Well, yeah, blogs are not academic journals. And if a blog post disagrees with academic consensus, the academic consensus will be right the vast majority of the time. But this applies to all blogs on the internet, not only those on LessWrong.

Those posts have grave errors, which the authors have been made aware of long ago but have refused to fix. For example, they literally made up the claim that wild animals have been getting more obese (there is no evidence of that). See this post I wrote for more details: https://www.lesswrong.com/posts/NRrbJJWnaSorrqvtZ/on-not-get...

I still think it's important to note that the blog's work is shoddy. The fact that this blog post is #1 on HN means that multiple people will read SMTM's other blog posts, which contain several falsehoods that they've refused to fix.

See, for example, this page https://manifold.markets/Natalia/how-many-of-these-falsemisl... for an incomplete list. I have been trying to get them to remove falsehoods from their blog for over one year, but it's impossible.

Note: I am the author of "It's Probably Not Lithium."

That SMTM post only addresses 5 of the ~17 studies about food lithium concentration that I mentioned in my post. It ignores several studies that did not use ICP-MS to measure lithium and found very low concentrations of it in food.

Moreover, as samatman mentioned, the post doesn't seem to understand that wet and dry weights are substantively different measures, and cites Magalhães et al. (1990) and Hullin, Kapel, and Drinkall (1969) as if they were reporting wet weight measurements of lithium, when they're reporting dry weight measurements. This makes a huge difference, because watercress (measured in Magalhães et al. (1990)) and lettuce (measured in Hullin, Kapel, and Drinkall (1969)) are both over 90% water by weight.

So that post is largely just based on cherry-picking and misrepresentation of studies.

Moreover, as far as I could tell, only *two* of their sources *actually attempt to estimate dietary lithium intake*, whereas almost all of my ~17 sources do. This is important because the SMTM post references a lot of studies that are not available on the internet, so we don't know whether they reported lithium concentration in dry or wet weight of food. Moreover, one of those sources based its estimation on hair lithium concentration rather than actual food measurements, using a regression model whose accuracy at low levels of dietary intake is unknown.

The vast majority of studies that actually attempt to estimate dietary lithium intake based on lithium concentration in food estimate it to be lower than 100 mcg/day, regardless of methodology. The only exceptions I am aware of are two studies from Manfred Anke.