Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

To Remember, the Brain Must Actively Forget

quantamagazine.org

Post reply on HN
Print view View original

111–120 of 126 posts

Re: To Remember, the Brain Must Actively Forget

#111

Post by ehnto »

2muchcoffeeman wrote: post #95
mattmanser wrote: post #82

Earlier quoted context omitted.

I'm not the grandparent, but after doing "Learning How To Learn" course on coursera[1] (very good), I finally realised I suffer from Aphantasia, where I don't see images in my imagination. Even though I'd tried loads of memory techniques over the years as I have a very poor memory, I'd never really understood people actually see images. I thought they meant that they were, like me, simply listing the properties the i…

I’ve never heard of Aphantasia before. Funnily enough it never occurred to me to think of things as pictures for my whole childhood. Even though I am almost a completely visual thinker. It’s worth thinking about how you think. Eg if I imagine time passing as images of numbers, I can keep track of time and talk at the same time. If I count them out in my head, I always loose track within seconds.

Oh wow, I just tried counting to ten with pictures while thinking a conversation. Neat trick!

I do something similar for separating my hands on the piano, the left hand becomes a pattern rather than literal chords and I can picture the hand placement and still think of the melody. Forever a work in progress that is though.

Re: To Remember, the Brain Must Actively Forget

#112

Post by Retra »

vanderZwan wrote: post #110
foxyv wrote: post #109

Earlier quoted context omitted.

I think a significant portion of my brain (Similar in size to the knowledge from my Bachelor's degree) was dedicated to remembering every quest and map up to WotLK. It's surprising how much I still remember from 7 years ago.

Funny enough, that might also reveal something about how human brains store information: locations on a map sounds suspiciously compatible with the method of loci (aka "memory palace", thanks to a more recent incarnation of Sherlock Holmes) [0] https://en.wikipedia.org/wiki/Method_of_loci

Music is also very easy to remember.

Re: To Remember, the Brain Must Actively Forget

#113

Post by Retra »

vanderZwan wrote: post #104
jeffwass wrote: post #15

Relevant convo between Sherlock Holmes and John Watson (My choice of Italics highlights). Below excerpt from A Study in Scarlet , Sir Arthur Conan Doyle, 1887. —— My surprise reached a climax, however, when I found incidentally that he was ignorant of the Copernican Theory and of the composition of the Solar System. That any civilized human being in this nineteenth century should not be aware that the earth travelled…

The flaw in the attitude of Sherlock Holmes is that he does not consider the, for the lack of a better term, "compressibility" of knowledge. Then again, Shannon wouldn't do his thing for another half century, so it would not be fair towards Sir Doyle to expect him to have taken this into consideration. The brain actually does a lot of lossy compression tricks to remember things together, plus I doubt that knowledge a…

I think there's a related problem that would have been understandable at the time: if you don't understand how the world works, you are potentially missing out on patterned information which allows one to form more accurate models of the world across domains. The assumption that's made seems to be that different problem domains are all independent, but that's not completely true. That's probably the main reason that mathematics and language are so useful.

Re: To Remember, the Brain Must Actively Forget

#115

Post by bizkitgto »

Aardwolf wrote: post #10

> Eleven of those genes, they discovered, were still active on one side of the animals’ brains but not on the other, even after the animals had apparently forgotten about the shock. What exactly does active genes mean here and how can they do that in a living slug and be different on both sides? Afaik, genes are mostly constant throughout one individuals' lifetime, and usually the same in all cells With almost a GB o…

Use it, or lose it!

Re: To Remember, the Brain Must Actively Forget

#116

Post by lupin3ken »

sirclueless wrote: post #90
mjburgess wrote: post #84

Earlier quoted context omitted.

The brain will keep around information that is used automatically . In my experience its a very bad idea to forgo learning something out of the expectation that it won't be used. I did this often, and it was catastrophic for me -- it turns out we're not very good assessors of what we will need. Doyle was a crank.

There's a related corollary to this question though, which is whether systems that require less human memory enable you to remember other things better. For example, your phone can dial dozens or hundreds of your friend's phones without you remembering their numbers. This means you never need to use their phone numbers and hence don't remember them. A reasonable question you might ask is whether you are now free to s…

So from what I understand, there is a tradeoff to whether or not you memorize something.

When you have a piece of knowledge memorized, it has the capability to be activated in parallel with other known knowledge. These parallel activations result in a new / novel idea to your neural network. If these activations haven't been had before, they may be a breakthrough in science, technology, or otherwise.

So in one aspect, if you do not memorize, you can not associate and create new things.

Luckily, with phone numbers and locations it might not be important or valuable to have them memorized. But that's hard to say. Many scientific breakthroughs and discoveries were made by people who had cross-discipline knowledge and associated two things that no one thought were valuable together. Potentially, even googleable factoids could potentially contribute to a scientific breakthrough.

Re: To Remember, the Brain Must Actively Forget

#117

Post by ZainRiz »

Waterluvian wrote: post #53
ZainRiz wrote: post #31

Earlier quoted context omitted.

There might be a market for this with high end consultants or psychiatrists as a more secure alternative to signing an NDA. It'll be something similar to the movie Paycheck You give the consultant the drug, then proceed to share your secrets with them and get their expert advice on your issue. Then by the end of the day/week/whatever they'll forget everything you told them and your secrets will be secure.

I feel that fundamentally won't work. Experts like that become experts because they are exposed to years and years of clients. One can build useful abstractions without giving away secrets.

I'm sure those experts would only accept jobs like this a very small percentage of the time and only for very large amounts of $$$. Otherwise they're not just loosing out on their expertise, they're loosing out on life

Re: To Remember, the Brain Must Actively Forget

#118

Post by ZainRiz »

Cyphase wrote: post #43
ZainRiz wrote: post #31

Earlier quoted context omitted.

There might be a market for this with high end consultants or psychiatrists as a more secure alternative to signing an NDA. It'll be something similar to the movie Paycheck You give the consultant the drug, then proceed to share your secrets with them and get their expert advice on your issue. Then by the end of the day/week/whatever they'll forget everything you told them and your secrets will be secure.

Philip K. Dick explored this idea in his short story "Paycheck"[0], which was made into a movie[1]. Edit: Did you edit your comment? I'm pretty sure you hadn't mentioned Paycheck. No problem, just making sure I'm not going crazy. :) [0] https://en.wikipedia.org/wiki/Paycheck_(short_story) [1] https://en.wikipedia.org/wiki/Paycheck_(film)

Lol, no, no edits :)

Re: To Remember, the Brain Must Actively Forget

#119

Post by vanderZwan »

Retra wrote: post #112
vanderZwan wrote: post #110

Earlier quoted context omitted.

Funny enough, that might also reveal something about how human brains store information: locations on a map sounds suspiciously compatible with the method of loci (aka "memory palace", thanks to a more recent incarnation of Sherlock Holmes) [0] https://en.wikipedia.org/wiki/Method_of_loci

Music is also very easy to remember.

IIRC there is recent research confirming it is at least partially stored in the same are responsible for muscle memory, which supposedly explains quite a few things.

Re: To Remember, the Brain Must Actively Forget

#120

Post by ZeikJT »

vanderZwan wrote: post #119
Retra wrote: post #112

Earlier quoted context omitted.

Music is also very easy to remember.

IIRC there is recent research confirming it is at least partially stored in the same are responsible for muscle memory, which supposedly explains quite a few things.

If that is true, the obvious question is why is it stored there?
Post reply on HN