I consider all of the behaviors you describe as basically transform functions. In fact, I consider any decision maker a type of transform function where you have input data that is run through a transform function (such as a behavior-executor, utility-maximizer, weighted goal system, a human mind, etc.) and output data is generated (and in the case of humans sent to our muscles, organs, etc.). The reason I mention this is that trying to describe a human’s transform function (i.e., what people normally call their mind) as mostly a behavior-executor or just a utility-maximizer leads to problems. A human’s transform function is enormously complex and includes both behavior execution aspects and utility maximization aspects. I also find that attempts to describe a human’s transform function as ‘basically a __’ results in a subsequent failure to look at the actual transform function when trying to figure out how people will behave.
Duncan
If I understand this correctly your ‘AI’ is biased to do random things, but NOT as a function of its utility function. If that is correct then your ‘AI’ simple does random things (according to its non-utility bias) since its utility function has no influence on its actions.
One of the primary problems with the rationalists, humanists, atheist, skeptics, etc. is that there is no higher level organization and thus we tend to accomplish very little compared to most other organizations. I fully support efforts to fix this problem.
I’d like to “Hold Off on Proposing Solutions” or in this case hold off on advocating answers. I don’t have time to list all the important bits of data we should be considering or enumerate all the current hypotheses, but I think both would be quite valuable.
Some quick hypothesis:
-Mr. Hat & Cloak is Quirrellmort & responsible for Hermione’s ‘condition’
-Mr. Hat & Cloak is Lucious & responsible for Hermione’s ‘condition’
-Mr. Hat & Cloak is Voldemort, but not the Quirrell body.
-Mr. Hat & Cloak is Quirrellmort and trying to take out Hermione as Harry’s good side anchor.
-Mr. Hat & Cloak is NOT Quirrellmort.
-Mr. Hat & Cloak is Grindelwald
-Author has made massive continuity alterations many of which are unclear thus it follows that relying on continuity is difficult at best. This most severely impacts character personalities and motivations.
Some quick puzzle bits :
-Hermione thinks Draco and Snape are doing / plotting something evil and hates Draco.
-Hermione was subject to a “Groundhog Day Attack” (I’m not even sure what this in the context of this story).
-Hermione’s tone of voice was all satisfaction just before being arrested for allegedly trying to kill Draco.
-Draco has no idea what’s going on.
-Hermione defeated Draco’s spell which was supposed to be extremely hard to do.
-Black Hat and Cloak did something to Hermione’s mind / memory.
-Prof. Q. has some major plotting going.
-Prof. Q.’s goals are not very clear, but his speech in the dinning hall may have shed some light on it and is consistent w/ convo’s w/ Harry.
-Author went out of his way to ‘rationalize’ how Harry didn’t go talk to Hermione. Not entirely clear that’s consistent w/ Harry or Hermione.
Okay, that’s all I’ve got time for...
If the sorting hat has enough access / ability to one’s mind to sort children into their appropriate house then it seems entirely possible that it has enough access / ability to identify a false memory. The sorting hat is an extremely powerful artifact which implies that the false memory would have to be a significantly greater power for us to conclude at this point that it can remain hidden from the sorting hat.
The hat says specifically: “I can go ahead and tell you that there is definitely nothing like a ghost—mind, intelligence, memory, personality, or feelings—in your scar. Otherwise it would be participating in this conversation, being under my brim.” It says memory specifically. Both a false memory and a ‘scar memory’ could at this point be treated as ‘foreign’ to Hermione.
Are you referring to this slightly earlier quote: “Anyway, I have no idea whether or not you’ve been Obliviated. I’m looking at your thoughts as they form, not reading out your whole memory and analyzing it for inconsistencies in a fraction of second. I’m a hat, not a god.”? Here the hat says it cannot detect deleted memories or do the sophisticated analysis required to discover tell tale ‘inconsistencies’.
If that is the case then the hat didn’t actually say “it couldn’t tell if Harry had any false memories.” It said it couldn’t detect deleted memories and seems to imply that ‘sophisticated analysis’ of all of his memories for ‘inconsistencies’ would be required to do so. The false memory given to Hermione is at the forefront of her mind and doesn’t require the hat to scan her memories (though Hermione could replay memories of event for the hat presumably). In addition the false memory is entirely out of character with Hermione’s personality which is something the hat, at a minimum, should be able to verify. Considering the quote specifically addressing foreign memory, it seems entirely possibly the hat may immediately detect the false memory for what it is.
Eliezer Yudkowsky’s Author Notes, Chp. 81
This makes me worry that the actual chapter might’ve come as an anticlimax, especially with so many creative >suggestions that didn’t get used. I shall poll the Less Wrong discussants and see how they felt before I decide whether >to do this again. This was actually intended as a dry run for a later, serious “Solve this or the story ends sadly” puzzle – >like I used in Part 5 of my earlier story Three Worlds Collide—but I’ll have to take the current outcome into account when >deciding whether to go through with that.Let me argue that this chapter was in no way an anticlimax:
We had no way to know which solution Harry might have come up with or picked (I like the hat trick still even though I figured that was not likely the solution of choice).
Neither Harry nor Eliezer are omniscient
Harry was under a lot of time pressure and had less information to work with than the readers
There is a lot of ‘motivation’ to keep the story interesting which limits the available solution space (i.e., any solution that results in a terrible story is not really an option)
A lot of entertainment is derived from ‘watching’ HOW things play out. History, movies, fiction, etc. are interesting to me for this reason and not because the main character did everything flawlessly.
If Eliezer does want to do a puzzle plot piece I strongly recommend accounting for fact that many of us read this story to relax and do NOT exercise our full investigative powers on solving story problems as that involves a lot of boring (for me at least) mundane work when done properly.
How do you propose organizing a ‘master list’ of solutions, relevant plot pieces, etc. given the current forum format? Some people have made some lists, but they are often quickly buried beneath other comments. I’m also not familiar enough with how things work to know if a post can be edited days after it has been posted. One obvious solution is that a HPMOR reader who likes making webpages puts up wiki page for this. Can this be done on Lesswrong.com?
Long term caffeine tolerance can be problematic. To combat this problem, every 2-4 months I stop taking caffeine for about 2 weeks (carefully planned for less hectic weeks). In my experience and that of at least one other colleague this method significantly lowers and possibly removes the caffeine tolerance. Two people does not make a study, but if you need to combat caffeine tolerance it may be worth a try.
I look forward to the results of this study. Quite frankly, most soft science fields could use this sort of scrutiny. I’d also love to see how reproducible the studies done by medical doctors (as opposed to research scientists) are. Quite frankly, even the hard sciences have a lot of publications with problems, however, these erroneous results, especially if they are important to current topics of interest, are relatively quickly discovered since other labs often need to reproduce the results before moving forward.
I would add one caution. Failure to replicate an article’s results does not necessarily mean the results are wrong. It could simply mean the group trying to reproduce the results had any number of other problems.
I would bet this is totally impractical for most studies. In the medical sciences the cost is prohibitive and for many other studies you need permission to experiment on organisms (especially hard when humans or human tissues are involved). Perhaps it would be easier for some of the soft sciences, but even psychology studies often work with human subjects and that would require non-trivial approval.
If you actually look at the genome, we’ve got about 30,000 genes in here. Most of our 750 megabytes of DNA is repetitive and almost certainly junk, as best we understand it.
This is false. Just because we do not know what role a lot of DNA performs does not mean it is ‘almost certainly junk’. There is far more DNA that is critical than just the 30,000 gene coding regions. You also have: genetic switches, regulation of gene expression, transcription factor binding sites, operators, enhancers, splice sites, DNA packaging sites, etc. Even in cases where the DNA isn’t currently ‘in use’ that DNA may be critical to the ongoing stability of our genome over multiple generations or have other unknown functions.
If a section of DNA is serving a useful purpose, but would be just as useful if it was replaced with a random sequence of the same length, I think it’s fair to call it junk.
Unless this is a standard definition for describing DNA, I do not agree that such DNA is ‘junk’. If the DNA serves a purpose it is not junk. There was a time when it was believed (as many still do) that the nucleus was mostly a disorganized package of DNA and associated ‘stuff’. However, it is becoming increasing clear that it is highly structured and that structure is critical for proper cell regulation including epigenetics.
If it can be shown that outright removal of most of our DNA does not have adverse affects I would agree with the junk description. However, I am not aware that this has been shown in humans (or human cell lines at least).
In the context of “what is the minimal amount of information it takes to build a human brain,” I can agree that there is some amount of compressibility in our genome. However, our genome is a lot like spaghetti code where it is very hard to tell what individual bits do and what long range effects a change may have.
Do we know how much of the human genome can definitely be replaced with random code without problem?
In addition, do we know how much information is contained in the structure of a cell? You can’t just put the DNA of our genome in water and expect to get a brain. Our DNA resides in an enormously complex sea of nano machines and structures. You need some combination of both to get a brain.
Honestly, I think the important take away is that there are probably a number of deep or high level insights that we need to figure out. Whether it’s 75 mb, 750 mb, or a petabyte doesn’t really matter if most of that information just describes machine parts or functions (e.g., a screw, a bolt, a wheel, etc.). Simple components often take up a lot of information. Frankly, I think 1 mb containing 1000 deep insights at maximum compression would be far more difficult to comprehend than a petabyte containing loads of parts descriptions and only 10 deep insights.
With the understanding that I only have a few minutes to check for research data:
http://www.ncbi.nlm.nih.gov/pubmed/1801013
http://www.ncbi.nlm.nih.gov/pubmed/21298068 - “cognitive response … to light at levels as low as 40 lux, is blue-shifted”
I’m confused as to what the point of the gate keeper is. Let us assume (for the sake of argument) everything is ‘safe’ except the gate keeper who may be tricked/convinced/etc. into letting the AI out.
If the point of the gate keeper is to keep the AI in the box then why has the gate keeper been given the power to let the AI out? It would be trivial to include ‘AI DESTROYED’ functionality as part of the box.
If the gate keeper has been given the power to let the AI out then isn’t the FUNCTION of the gate keeper to decide whether to let the AI out or not?
Is the point simply to have a text communication with the AI? If this is the case why bother stipulating that the gate keeper can let the AI out. If humans can be subverted by text there is no need to utilize a built in gate it seems to me.
I think the CFAR is a great idea with tons of potential so I’m curious if there are any updates on how the meetup went and what sorts of things were suggested?
“1. Life is better than death. For any given finite lifespan, I’d prefer a longer one, at least within the bounds of numbers I can reasonably contemplate.”
Have you included estimates of possible negative utilities? One thing we can count on is that if you are revived you will certainly be at the mercy of whatever revived you. How do you estimate the probability that what wakes you will be friendly? Is the chance at eternal life worth the risk of eternal suffering?
I am having trouble scanning the HPMoR thread for topics I’m interested in due to both it’s length and the lack of a hierarchical organization by topic. I would appreciate any help with this problem since I do not want to make comments that are simple duplicates of previous comments I failed to notice. With that in mind, is there a discussion forum or some method to scan the HPMoR discussion thread that doesn’t involve a lot of effort? I have not found organizing comments by points to be useful in this respect.
Edit: I’m new and this is my 1st comment. I’ve read a lot of the sequences, but I don’t know my way around yet. It’s quite possible I’m missing a lot about how things work here.