@JhanicManifold's banner p

JhanicManifold


				

				

				
6 followers   follows 0 users  
joined 2022 September 04 20:29:00 UTC

				

User ID: 135

JhanicManifold


				
				
				

				
6 followers   follows 0 users   joined 2022 September 04 20:29:00 UTC

					

No bio...


					

User ID: 135

What GPT does is predict the next token. That's a simple statement with a great deal of complexity underlying it.

At least, that's the Outer Objective, it's the equivalent of saying that humans are maximising inclusive-genetic-fitness, which is false if you look at the inner planning process of most humans. And just like evolution has endowed us with motivations and goals which get close enough at maximising its objective in the ancestral environment, so is GPT-4 endowed with unknown goals and cognition which are pretty good at maximising the log probability it assigns to the next word, but not perfect.

GPT-4 is almost certainly not doing reasoning like "What is the most likely next word among the documents on the internet pre-2021 that the filtering process of the OpenAI team would have included in my dataset?", it probably has a bunch of heuristic "goals" that get close enough to maximising the objective, just like humans have heuristic goals like sex, power, social status that get close enough for the ancestral environment, but no explicit planning for lots of kids, and certainly no explicit planning for paying protein-synthesis labs to produce their DNA by the buckets.

This part never usually pans out lol.

Oh it did for me, I still remember her reaction when she saw me for the first time in like 6 months: * looks at me, does a double take, eyes widen, face becomes fully red, furtive looks the whole evening *. I actually feel a little bit guilty about just how good revenge feels.

Yeah the pulsing patterns seem very specific, and are probably the entire technical moat of the company. The way it works is that there are a variety of "programs" on the app, so you have a "stress program" that lasts 15 min, which starts with short, intense pulses that get quicker and quicker, then you have a "calm program", a "sleep program", etc. The device modulates both the intensity and the frequency of the haptics over time depending on the program you chose.

Get a meat thermometer and grill to 170F internal temperature. I mix 4 eggs, 2 cups panko bread crumbs, 1.3kg of extra lean beef, 3 tbsp smoked paprika, 2 tbsp garlic salt and 2 tbsp crushed oregano together into the burger paste, and they turn out great every time if grilled at around 450F. Spices don't burn if you mix them into the meat itself.

Short answer: it depends on how much cardio you're doing already. Cardio contributes to systemic fatigue, and too much of it will reduce your NEAT (Non-Exercise-Activity-Thermogenesis), basically you'll fidget less and be less inclined to take the stairs instead of the elevator, which will have a net-negative effect on caloric balance.

Very relevant video: Does More Cardio Equal More Weight Lost?

(and I'm very surprised by your 100cal per 2 miles walking number, I use this calculator for estimating walking calories, which gives me much higher numbers)

Pinging @JhanicManifold.

Lol, I guess I have been writing a lot about how to get various psychedelics, peptides, and semaglutide.

I should really make a giant "Nerd's Guide to Getting Ripped & Shredded" as a post that we could always link for these questions, rehashing everything every wednesday seems like a bit of a waste of time.

You got injured doing goddamn Kegels?! How is that even possible?

Not injuring yourself in the gym is pretty easy:

  1. warm up with 5 to 10 minutes of cardio

  2. Don't do weights close to your one-rep max

  3. Slowly increase your total weekly volume to give your tendons time to adjust

  4. Don't improvise in the gym, instead have a pre-planned routine with exact weights on an excel sheet that you stick to

  5. If you're doing complicated movement patterns like squats and deadlifts, make sure your form is correct

No, literally just free physical letters to everyone who requests it, like all banks do. Twitter should have more than enough money to do this for all US users.

I have, perhaps surprisingly, managed to make new close friends as a mid twenties guy. Here's the trick, you first say "hey bro (or girl I want as a friend), I really enjoyed talking to you at that thing that time, we should get coffee sometime, how's next thursday?". Next, you need to talk one-on-one, it's hard to make a new close friend when you're always in a group with them, and you need to (gradually) tell them some of your secrets, vulnerability is required for close friendship. Virtually everyone to whom you offer one of your small secrets will reciprocate with one of their own, and thus the friendship builds.

Then, once you've hung out a few times one on one, you hit them with a heartfelt "hey man, great hanging out today, this is kinda weird to say, but I want you to know that I'm really grateful to have you as a friend". Think of yourself, would you allow yourself to grow closer in friendship with someone you weren't sure wanted the same thing? Realising that someone you considered a close friend doesn't feel the same probably hurts at the same level as romantic rejection. That's why the earnest declaration that you consider them a close friend works so well, it alleviates their fear that their friendship will not be reciprocated. People do want close friends, they're just afraid to hope.

Ben Shapiro also makes the point that they drop "knock bombs" before the real bombs, the only purpose of those bombs is to shake the building to tell civillians to evacuate.

Buying it on indiamart.com from India pharmacies, I haven't really had problems with importing it in Canada

I might have fucked up one of the easier ones, but gotten avulse correctly. That would explain things if difficult questions are worth more.

hmm, top 0.2% for me, I'm kind of surprised, I never made any explicit effort in all my english classes, and I'm not even a native speaker.

/images/1685626469587057.webp

Athens was filthy. Almost every building is completely covered in bad graffiti up to a certain height.

And don't forget the brothels... they're called "studios" there, and there was a truly surprising number of them when I went last summer.

The recommended dosing has you slowly ramp up the dose over like 2 months until you get to the truly effective amounts. I suspect that you would get very large gastro-intestinal side-effects if you just went from 0 to the effective dose. Semaglutide also has a very long half-life of a week, so if you're injecting 1.0mg/week steadily, to get the same blood concentration with a single injection you'd need 2.0mg. I think it would be a fairly bad idea, there's a good chance that the diarrhea and vomiting you'd get would wipe out any mental benefits from the lack of hunger.

Hmm, somehow I doubt that. It's not like all schoolgirls I pass by giggle, but it's happened like 3 times over the past month (I probably pass by more schoolgirls than you because I take the subway). But then I'm curious, what things did you notice in your case?

yep, got a 25.38 score, which makes me HighRat too

I think instead of reading the documentation of libraries I'd try to read the main books on the theory, but only reading the first part of every chapter and very lightly skimming the rest, doing none of the exercises. Like, you understand what an integral is within like 30 minutes of having the problem stated to you, and the best ressources to be introduced to the problem are still the textbooks, they just happen to come with hundreds of pages you don't need. When I first learned ML I just spent 2 days reading Murphy's 2012 book without worrying about any of the details, I just wanted to get a small introduction to literally every method so that I could get a sort of mental picture of the entire field.

The problem with only knowing the theory so shallowly is that you're kind of brittle to modifications and expansions of the problem statement. Like, how would you solve the following problem: you have a company that has an industrial process with 10 free parameters, and they have a yield function F(x) : R^10 -> R that they want to maximize, but the exact physics if the process is unknown or very complicated. Each trial run to compute F(x) costs 1 million dollars (and of course F(x) has some unknown variance), so so far they only have around 30 data points {x_i, F(x_i)}. Your job is to advise them on how to pick the next point x_i to try in order to maximize their profits (which increase with yield, but decrease with additional suboptimal trials). And what if F(x) is nonstationary? Maybe the machines degrade over time, and the optimal parameters change...

Unfortunately, yes, small amounts of alcohol have a detectable effect on brain white and gray matter volume:

Specifically, alcohol intake is negatively associated with global brain volume measures, regional gray matter volumes, and white matter microstructure. Here, we show that the negative associations between alcohol intake and brain macrostructure and microstructure are already apparent in individuals consuming an average of only one to two daily alcohol units, and become stronger as alcohol intake increases.

Here "one unit" means 10ml of ethanol, slivovitz has 50% abv, so you're drinking 350ml ethanol per 14 days = 2.5 alcohol units per day. The paper I linked has a bunch of interesting figures (fig 3 in particular is nice), and they provide this useful comparison:

For illustration, the effect associated with a change from one to two daily alcohol units is equivalent to the effect of aging 2 years (or 1.7 years in the model that excludes individuals who consume a high level of alcohol), where the increase from two to three daily units is equivalent to aging 3.5 years (or 2.9 years in the model that excludes individuals who consume a high level of alcohol).

Going from 0 to 1 daily units doesn't have any measurable effect in that study, but going from 1 to 2 and 2 to 3 does. So with your 2.5 units/day, it's equivalent to an aging-related decrease in brain volume of around 3.75 years. Not world-ending in any sense, but still not nothing.

(I'm slightly confused by the study, since presumably the effects should depend on how long you've been drinking alcohol, and not just provide a flat decrease in brain matter, but I'm not seeing any such effect reported in the paper)

I have a very large google sheets document where I compute macros per weight for a bunch of my most common homemade foods (like protein bars, chili, hamburgers, fried chicken, etc.), like the picture shows. Some variety does need to be sacrificed, but I arrived at a combination of 5 or 6 recipes that I like and that fit my macros pretty well.

/images/16756394291295397.webp

edit: also, I get cheap whey here and cheap casein protein here, direct import from india from indiamart would be even cheaper, but that's a bit too much hassle for me right now

modafinil was generally for wakefulness?

Just like caffeine both keeps you awake and focuses/motivates you, modafinil also has quite significant effects on focus and motivation, it's sort of the intermediate step between caffeine and amphetamines. I wouldn't necessarily use it everyday, otherwise you get used to it, but it could really give you 2 days per week of great motivation.

Oh, very true! I honestly forget that people do these recreationally and not to understand the nature of self and sensory experience. Though I personally find it hard to think of anything less pleasant than doing shrooms in public.

yeah I just use insulin needles from amazon, with alcohol prep pads. The only thing you can't get from amazon is bacteriostatic water, which is still over-the-counter, just not from amazon (I got it from here). Peptides in general are best stored dry in a freezer or fridge, but I keep them at room temperature away from the light, the tests I've seen don't really show any meaningful degradation in a few months. Though they start to degrade much faster once you add in water.

Another thing, you probably won't be able to just buy a month's supply, the minimum order quantity is 10 vials of 2mg, which comes in a prepackaged little box disguised as a chinese beauty mask. I don't think the supplier is set up to ship orders which aren't multiples of 10 vials.

Surprisingly, it kind of does! It felt to me like it helped me not think of work or other things while I'm going to sleep, the vibrations on your skin have a way of capturing attention very effectively. I tried it out for a few weeks after seeing it recommended here, but I'm now returning it, the difference just isn't that big for me, nowhere near the magnitude that the bedjet is making.