Context: * One way to measure how good LLMs are which is gaining traction and validity is the following: * Let T(t) be the time it takes for a human to the task t. * Empirically, a given LLM will be able to fully accomplish (without extra human correction) almost...
This is Part I in a series on easy weightloss without any need for will power. The Origin: listening to the dark corners of the internet Losing weight is supposed to be really hard and require a lot of willpower according to conventional wisdom. It turns out that it was...
This is coherent with my experience. I'm pretty sure there are other problems solved by self-deception other than hostile telepaths. One other such problems solved by self-deception which I'm pretty sure I've seen in people is preserving motivation: if something is really important for me and I need to put in a lot of effort to make it happen and probability of success is very low (let's say epsilon), and if know that the probability of success is epsilon would totally annihilate my motivation to work towards it, then maybe hiding to myself that low probability safeguards my motivation to put in all the work necessary.
Think here at a situation where there is an natural catastrophe, and someone's loved one is caught below the rubble, the person refusing to believe that the person might be dead and doing everything to remove the rubble as fast as possible.
Maybe this is also where the planning fallacy comes from in some cases.
Yes. That is still planned!!! I'm just very bad at writing.
Primer: I've been collecting more data since and something super weird happened. I tried to gain more weight again to redo experiments, it was suprisingly harder than expected to gain more weight, but I managed. But super weird. After gaining more weight, going back on the half-assed potato diet didn't work as well anymore. I still didn't manage to loose the weight I intentionally gained! If I went on a total potato I would loose weight. But the semi-potato diet is not enough to compensate the days where I go to a party (or there is some sort of event with lots of food)! Super interesting because previously the semi-potato diet could easy compensate for those big meals days.
Thanks for this info. Ya this really goes in the direction of what I think is happening.
Not really. It's an ion. Your body easily eliminates anything which is water soluble in your pee.
Ya, all that sounds about right to me :) Thanks for writing out so clearly :)
I totally believe that a low potassium 500 kcal diet would see rapid and significant weight loss. My experience so far tells me that I would expect doing a 500 kcal diet on low K would be very difficult (my body would just painfully crave food) whereas with high K it would make it much easier.
Wow! Thanks for all the detail. You seem to have a precise and detailed knowledge of how your body works! I'm impressed.
I did it at the belly button, but I did it at lungs-full because I thought it would be harder for me to cheat myself at lungs full. lungs neutral felt like I could unconsciously be little less full when it would support my hypothesis and little more full when it wouldn't ...
Oh wow!! Great data! Thanks for that.
So my incomplete tests for the moment seem to indicate that if I take no potassium and no calorie-not-dense meal, then I gain weight. If I just take ~2500 mg K or more but no calorie-not-dense meal, I lose weight very slowly, if I just take one calorie-not-dense meal a day but no K I lose weight very slowly, but if I do both, then I lose weight visibly. Do you think something like that could be consistent with your experience?
This is Part I in a series on easy weightloss without any need for will power.
Losing weight is supposed to be really hard and require a lot of willpower according to conventional wisdom. It turns out that it was actually really easy for me to go from a BMI of above 29 (30 is officially obese) to below 25 (normal is 18 to 25) in 3½ months. And knowing what I know now, I think I could easily do it again in a 1½ month.
I'm not someone who ever tried dieting before. Dieting sounded like a lot of effort and willpower for very... (read 2543 more words →)
The doubling time is not 53 days but roughly 6-7 months.
https://www.lesswrong.com/posts/GHxWxRFto32jedmJh/prediction-recursively-self-improving-ai-for-2033
From your own METR source: