All of Dentin's Comments + Replies

My belief is that it's primarily the voting system that causes this. (Not the electoral college; rather the whole 'first past the post' style of voting.) We see scissors presidents because that's the winning strategy.

I suspect that other more sophisticated voting systems (even just ranked choice!) would do better. No voting system is perfect, but 'first past the post' is particularly pathological.

They are hard to fix, but you quickly realize that they're all one-off problems and whatever solution you come up with for one person isn't going to work for anyone else.

If you can even find a solution.

The problems are too hard, the feedback is too weak, and the solutions never generalize. All of those are the opposite of what I'd prefer out of my problems.

I do the majority of these, and converged on them independently over the course of decades. I ended up doing most of these because they made my life better.

Complete newbie question: is it possible to construct a version of these models that uses a 3 dimensional vector, instead of the 768 dimensional vector?

From the sound of it, the 768 dimensional vector is basically a constant linear transform of the three PCA components. Can we just declare the linear transform to be a constant array, and only train up the three components that appear to be the most needed? Eg generate the 768 from the PCA?

My personal bloodwork did not. It shows effectively no change over the past two and a half years - including zero change between late last year (when I was having serious difficulties), and last month (when I had been supplementing protein by about a hundred grams per day.)

So I don't know what that test is measuring, but it doesn't appear to be related to protein levels.

(It's also worth noting that not all protein is created equal. I had a fair amount of plant protein in my diet, but the amino acid profile used to make plant cells is not the same as the the amino acid profile needed to build muscle and repair sinew. I believe I was short on specific key amino acids.)

Congrats! I went through this thought process as well, and one of your three hypotheses above seems like the right one. Vitamin D isn't the issue (I have tests for it and have heavily supplemented for years), and sulfur itself isn't an issue (onions and broccolo are both pretty big in my diet). However, the lack of sulfur amino acids is the lead hypothesis.

Over the years, I had slowly shifted my diet more and more plant based: lots of vegetables, with occasional meat and a piece of fish every couple of days. As you mentioned, not all protein is create... (read more)

Funny you should mention this; it made me check my records. It turns out that none of the doctors actually requested bloodwork. However, I do have my own bloodwork, which I do every 4-6 months on my own. Looking through that, what I see for heptatic protein level is 7.2+-0.2 for the past two and a half years.

This includes my most recent test, where I had been taking massive amounts of protein for months. So whatever that test is measuring, it doesn't actually seem related to the amount of protein the body has available or needs.

Yeah, a big part of the problem was expert ping-pong. I only saw one doctor twice. It's part of the global dysfunction I observed. Because of that, I have been seriously considering signing up for concierge medicine, which has a similar business model to what you describe above.

FYI, structural conformation diseases are actually quite common in humans, we just call them something different: amyloidoses. For example, TTR amyloidosis kills a good fraction of our oldest humans.

From https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2634591:

"Prions were originally defined as a unique class of infectious agents, whose infectivity relates solely to protein. In mammals, they cause fatal neurodegenerative diseases, such as Creutzfeldt-Jacob disease of man, sheep scrapie and bovine spongiform encephalopathy (reviewed in refs. 1 and 2). All the... (read more)

Yep, that pretty much handles it. Thanks for the update!

I don't find it surprising; 0.1% is a fairly low bar here on LW. I'm not considered that unusual here, and my calibrated guess is that I'm in the 0.3% category. There's a million people in the USA alone at that level, and three hundred thousand at 0.1%. That's a wide pool to select from.

Personally I wouldn't be surprised if Musk was substantially above the top 0.1%. I've seen a number of technical interviews with him; he and I have similar backgrounds and technical field strengths; and we are approximately the same age. I feel able to properly evaluate his competence, and I do not find it lacking.

For a really good example of what I would consider a 'dumb' way for AGI misalignment to be problematic, I recommend "accelerando" by charles stross. It's available in text/html form for free from his web site. Even now, after 20 years, it's still very full of ideas.

(FYI, sections are about ten years apart in the book, but last I read it it seemed like the dates are off by a factor of two or so. Eg. 2010 in the book corresponds loosely to 2020 in real life, 2020 in the book corresponds loosely to 2040, etc.)

In that book, the badness largely comes from increasingly competent / sentient corporate management and legal software.

I totally understand where you're coming from, and I apologize for straddling the line on Norm One. I saw that it was heavily downvoted without comment, and that struck me as unhelpful.

Regarding the post itself, it wasn't a matter of being unable to understand. It seemed likely to me that there were insights there, and that if I spent enough time on it I could pull them out. It was more about the other demands on my time, which I suspect isn't a unique circumstance.

Regarding probability of mistake, I think that's an unhelpful way of looking at it. IMO ... (read more)

Real quick, minor readability concern: I was about a quarter of the way through the post, and fairly confused, before I figured out that the line in the pictures was facing backwards from my mental model.

IOW it wasn't intuitively obvious to me early in the post what 'in front of' or 'behind' meant. It might be worth indicating front/back on the images.

Not downvoting as I see some potential here, but also not upvoting. This post is very long, with little structure, and an unclear / tedious to derive takeaway. I'd recommend at a minimum splitting into sections, such as the AstralCodexTen "I II III IV V VI ..." scheme with opening/closing summaries. I would also guess at least half of it could be removed without damaging what you intend to convey.

In other words, there might be good content / ideas in here, but it would take too much effort for me to extract them. There are a great many things competing for my time, and I must be choosy about what I spend that time on.

AIUI, you've got the definition of a p-zombie wrong in a way that's probably misleading you. Let me restate the above:

"something that is externally indistinguishable from an entity that experiences things, but internally does not actually experience things"

The whole p-zombie thing hinges on what it means to "experience something", not whether or not something "has experience".

To be fair, even if what you're referring to above is true (I don't believe it is - lookup table compression is a thing), it's an implementation detail. It doesn't matter that a naive implementation might not fit in our current observable universe; it need merely be able to exist in some universe for the argument to hold.

And in a way, this is my core problem with Searle's argument. I believe you can fully emulate a human with both sufficiently large lookup tables, and also with pretty small lookup tables combined with some table expansion/generation code... (read more)

Please clarify/reword your statement; I can't figure out what you're trying to say. The word "that" is almost completely unspecified.

I'd go way more limited:

• No elected person may hold another elected position in any branch of government for at least one year and one day after the last day of their current term, even if they do not complete their term.

• No branch of government may have direct control over the parameters or structure of elections or appointments for any position within the branch. This includes districting, type of voting system, timing, and election rules. If elections are for all branches, an independent party must be responsible for elections.

• All elected and a

So much yes to this post. This tracks with about everything I've experienced so far. It also makes me appreciate even more the close friends I have in the medical profession; I know I can trust them to review ideas, think about them, theorize, and suggest other areas for research at a level that is appropriate for my skillset. In private, they admit to proposing numerous luck based treatments that panned out. Our hardware is complicated, and we have extremely limited monitoring and visibility on it. Doctors that aren't burned out and are curious know th... (read more)

I agree. I've been donating $10k-50k per year for the past decade or so. I determined a couple years ago that it was better for me to acquire money at my current job and spend it hiring professionals, than to go into fundamental research myself.

Most of my hobby time these days goes toward biochem and biomedical research, so that I can be at the cutting edge if it becomes necessary. Being able to get treatments from 5-10 years beyond the official approval timelines may very well make the difference between life and death.

I've had enough time and exposure that I've largely worked through my fear of death; I put substantial effort into finding a healthy way of managing my mental state in that areas. It doesn't significantly impact my day, and hasn't for a while.

But it's still there looming large when I ask myself, "what's the most important thing I can be doing right now?"

Your question resonates strongly with me.  About ten years ago, I decided to try Uberman to solve exactly this problem.  While it didn't solve the problem, it taught me a lot about sleep, and it was worth three weeks of hell for that alone.

In the time since then, I've come to the conclusion that it's probably not that sleep recharges our available mental energy.  Recharging is part of it, but I think a bigger factor is probably the fact that sleep provides memory consolidation, and most importantly, sleep makes us forget things.  We are... (read more)

I would prefer plain text, or at least dramatically more compact.  I find glowfic, including this one, to be borderline unreadable because of the format.

Like most delusions there's some truth here.  But only some:

 - You're not a "human superintelligence running on the substrate on this human's brain."  You're an ordinary human intelligence running on the substrate of a human brain.

 - You're not fully aligned with human values, you're aligned with the values that you think are human values.

 - The "voice of Humanity and Rationality and Truth" may call out to you deeply, but many things call out deeply to many people.  For some, it's Jesus, for others it's the healing power of cr... (read more)

Yeah.  The current Doomsday Clock is what happens when an organization has outlived its usefulness and is trying to remain relevant even though it really isn't.

I spent a bit of time looking at Geert Vanden Bossche's ideas about six months ago.  I came away extremely unimpressed; he takes reasonable sounding things ("viruses mutate under selective pressure") and tries to extrapolate from them things that we don't actually see in reality.  He describes a plausible sounding reality, that is not our reality.

I find his post incredibly uncompelling.  I believe he's arguing against a straw man; economists in the real world doing real work involving real dollars are painfully aware of the issues he brings up.

My guess is that his experience in economics is heavily influenced by his PhD work and that he's arguing against "economics as he experienced it at universities" as opposed to "economics as practiced by professionals in a real economy".

https://blogs.sciencemag.org/pipeline/archives/2021/06/07/ivermectin-as-a-covid-19-therapy is one of the better, current overviews of Ivermectin.  Basically, we don't have enough information.

One other thing to consider is that even small differences in replication rate might actually matter.  Consider that it takes a week for the virus to really ramp up, and that's a large number of doubling periods.  Even just getting a larger or smaller initial dose seems linked to how sick people get.  Even a few percent difference may allow the immune system to stay ahead in the arms race, and result in a nonlinear change in death rate.

Note that I'm not saying this happens; I'm saying that because this is an exponential growth attacker (the virus) versus and exponential growth responder (the immune system), even small differences in growth rates might have a large impact.

Yeah, his second claim is bogus.  That's not how it works, and that's not what we've been seeing with existing mutations.

As an example, look at E484K - this mutation changes the amino acid polarity, so that antibodies trained against the E variant will have a much harder time attaching to the K variant.  If an antibody fails to attach, it doesn't 'crowd out' anything.

In the case where an antibody attaches but doesn't actually "inactivate" the virus due to a mutation, that's because the virus' attack surfaces are still present and exposed (otherwi... (read more)

#1 is where I would hinge a lot of objection.  Specifically:

"The vaccines are targeting outdated variants, and some vaccines are already only partly efficient. This creates the perfect conditions for further viral evolution."

Yes, the vaccines are targeting outdated variants, and yes, the vaccines are only partially efficient.  But the mRNA vaccines, even partially efficient, are still hugely overkill.  From previous posts here on LW, even partially effective mRNA vaccines likely cut transmission by a factor of 100 between the reduced infecti... (read more)

These extremely short responses discarding the bulk of my content feel less like you're attempting to understand, and more like you're attempting to get me to draw bright lines on a space I have repeatedly indicated is many different shades of grey.  Disconnecting from the discussion for now.

Publisher, advertiser, the distinction does not matter.  The point is that the target does not get to decide.

You might find it unpleasant, but it's it the job of Simurgh Followers to spread the Truth Of The Endbringers to everyone!  Surely if people just watch enough of it, they will be converted.

The point is that the target gets to decide what's acceptable and what isn't, not the advertiser. The current system makes the advertiser the judge, and that's not ok, even if we have managed to construct a sorta functional system that mostly takes care of the worst abuses.

I'm not convinced I fully understand your distinction, let alone that we could codify it sufficiently to make it into law.

Regarding 'codify into law', that's not an excuse, and it disregards how the US legal system works.  If we can codify slander, if we can codify "harm", if current advertising companies can codify "unacceptable ad", we can codify this.

If you visit a model railroading site, are ads for model locomotives push or pull?

Firm push, but only because of the physical realities of the current system.

The fact of the matter is that by default, ... (read more)

For bulk exchange of information and state, holistic is really good.  I strongly prefer the holistic approach, but I've found that it only works for entirely friendly conversations.  If it's adversarial, I find that branches get aggressively pruned to just the things that the opposing side can most easily attack.

And if you think about holistic being optimized for "exchange of information and state", this makes perfect sense:  adversarial conversations are rarely if ever about information exchange; they're about "winning".

It's also perhaps wo... (read more)

Given the current vaccination rates in the US, and the fact that supply is already beginning to exceed demand, I'd recommend full open in approximately a month.  That gives most of the remaining unvaccinated people time to go get at least their first shot, and should allow us to get below exponential growth nationwide, even though we're likely to have it in sub-populations.  IMO the target should be 'not overloading hospitals'.

After that, let it burn.

From the interviews and things I've seen so far, literally the only way to change a vaccine denie... (read more)

I'd just like to point out that while "facing these tradeoffs is stupid and avoidable" (which I agree with), it's much, much more accurate to say instead "facing these tradeoffs is effectively impossible to avoid even though it's stupid and avoidable".  We might not like reality, but it's not going to go away no matter how much we call it stupid and avoidable.

Having an overarching model (or several competing models) of which different parts can be tested independently seems like a structure which is very amenable towards different scientists, so I am disappointed none of the biological/medical community has started doing something like this.

This is actually done, quite a lot in fact, it's just really hard and the search space is huge.  Kudos to you for your analysis; it's unlikely to be a major step forward, but given that idea search space is effectively exponential, it's also entirely possible that it's ... (read more)

> its entire purpose is to alter people's mental state without their permission

I think that's the core of our disagreement?

Yes, and I think that would be a better path to attack my position.  There's two attack vectors in that quoted line - "alter peoples mental state without their permission", and "permission".  I would recommend avoiding the first attack vector; that will be an exceedingly difficult sell to me.

Permission on the other hand is already a partially open attack vector, and you're much, much more likely to change my mind by that r... (read more)

Here's another version of your example:  Some people aren't watching nearly enough snuff and torture videos. There are people who would like to watch them, but don't know it exists.  If I place ads for torture and snuff videos and some people decide to click on them while other people don't, is that a problem?

As I mentioned earlier, advertising is like weaponry.  Your example also reads to me like a classic justification for 'everyone having guns':  "but what if I'm attacked by a rabid dog?  If I have my gun I can protect myself! &... (read more)

As someone who also works on Ads at Google, I have to take the opposite stance; I view advertising as a blight upon the face of humanity, something to destroy if we can at all figure out how to do so.  I comfort myself knowing that Google Ads is arguably the best of what's an awful ecosystem, and that I work in what's arguably the 'least bad part of advertising', which is fraud and abuse protection.  At least the systems I work on make things less terrible.

However, the 'least bad part of advertising' is still not 'good'.

My favorite analogy for ad... (read more)

I still have some remaining bitcoin, from the olden days when mortal man could mine it themselves.  My advice to everyone I've ever talked to regarding bitcoin is to avoid it.  I have been slowly divesting my holdings.

My rationale is that while both the dollar and bitcoin are fiat currencies, bitcoin is far, far less anchored to reality than most 'normal' currencies.  The dollar and the euro at least have people trying to keep monetary levels somewhat tied to physical economic value.  The value of bitcoin, meanwhile is largely driven by... (read more)

I've found lighting, melatonin, and caffiene regulation to be wonderful additions to my sleep regime.  I take melatonin pretty consistently at around 8:30 pm, and it seems like it helps make me sleepy ~45 minutes later.  As per SSC though, melatonin isn't particularly strong and the effect I'm noticing may very well be placebo.  I always have caffiene, but rarely after 2 pm, and typically not more than two cups of coffee per day.

That said, I suspect my lighting and light policy is having a much, much bigger effect.

The primary light source in... (read more)

Quick comment:  I noticed that in all of your examples above, I chunk substantially bigger and fewer pieces.  For example, in the "15 different bold bits" clip, I chunk it into about 8 pieces instead.

This is likely experience/background dependent; I happen to have a relatively strong background in ML and have read a stack of research papers recently, so I probably have both stronger noise filters and more complicated primitives available.

One possibly interesting side note:  I never once, in any of your examples, considered metadata about the... (read more)