Scientists Claim Brain Memory Code Cracked - 6 bit bytes | Hacker News Reader

Scientists Claim Brain Memory Code Cracked - 6 bit bytes

sciencedaily.com

94 pointsedderly39 comments

Once again misleading title. They did some modeling that shows that it might be possible to use CaMK2 to store information on microtubule lattices. That's purely simulations that show possible interactions between two proteins. No experiments were performed and they did not set out to break the brain code. There are no experiments that show information encoding in microtubules. CaMK2 is known to be involved in plasticity (in spine shape modifications via actin dynamics), but microtubules are assumed to be in the dendritic shafts only thus isolated from phosphorylated CaMK2. It would be worth to investigate further though.

As is often the case, a Reddit comment is more illuminating than the article:

http://www.reddit.com/r/science/comments/qqg5y/scientists_cl...

FTFA: Hameroff, senior author on the study, said "...We MAY have a GLIMPSE of the brain's biomolecular code for memory."

I know Hacker News prefers the original headline, but there has to be an exception for ScienceDaily, PhysOrg, and any other blog that tries to write linkbait summaries of peer-reviewed research articles.

This is the dingding step of the eternally true http://www.phdcomics.com/comics.php?f=1174

Extremely exciting! It's in PLoS, so the paper should be freely available as well.

Interesting that it's 6 dimensional, this figure shows up often in the brain: http://en.wikipedia.org/wiki/Grid_cell

The full article text is at http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fj... (yay PLoS!) -- read the inline Abstract and Author Summary to get a better description than the sciencedaily spam.

Sadly, nobody can crack the thought process of whoever designed sciencedaily.com (three equal width columns? 10% content and 90% blogvomit?).

blogvomit is a good word.

Why hasn't sciencedaily.com or physorg been banned from HN yet? Anybody know?

Incidentally, I've flagged so many physorg submissions that the flag link has been disabled for my account.

If it's any consolation, my flag link vanished a long time ago after I went on a "purge all the self-promotional infidels" jihad last year.

I consider it a badge of honor to have restricted features but still have a working account. They care enough to limit us, but they don't want us completely gone.

I'm torn on the subject of sciencedaily. Without the linkbait-y headline, I doubt this paper would have any attention here. They should turn into a HN-like site with links to articles, then give each link a short (under 100 words) lay person description not just copy/pasted from the article (maybe even with linkbait conclusions). Sometimes you have to bend the truth (i.e. marketing) if you want people to get excited about your work.

>I consider it a badge of honor to have restricted features but still have a working account. They care enough to limit us, but they don't want us completely gone.

Hey, that's a nice way of looking at it.. I lost my flag link during "bitcoin week" last summer.

I lost mine recently too. I'd been trying to keep the number of political stories down as they very rarely have any useful discussion.

Same! Apparently I took the article submission guidelines too seriously?

I guess loss of flagging rights might be due to flagging highly upvoted articles. Maybe searching for less popular articles from these spammy sources, and flagging those, would be a viable strategy.

Ever since I upgraded to Lion, a double tab in Safari to zoom into the content part of the text is all it took to avoid the vomit.

The headline of the submission seems misleading. It sounds like scientists have theorized a plausible model for how the brain stores information, not that they actually tested this in vivo.

The most interesting bit for me is the estimate for total capacity, which is ~10^19 bits.

That's about 2^63 bits, 2^60 8-bit bytes, 1 exabyte, or (probably) the sum of all Google data centers, in one brain.

Which means in a couple decades it will fit on a thumbdrive.

Except it will actually be a thumb drive

FTA: [S]ynaptic components are short-lived while memories last lifetimes. This suggests synaptic information is encoded and hard-wired at a deeper, finer-grained molecular scale.

It suggests nothing of the sort. It just as easily could mean that memories are stored at a shallower, coarser-grained scale. Raindrops are short-lived, yet rivers last thousands of years. Does that mean we can't model hydrodynamics without accounting for quantum mechanics in the H2O molecule?

Wow! Unclear if this is fact but 2 things surprised me.

- I wasn't convinced it would be base 2.

- I thought if it was, base 8 made more intuitive sense.

The former mistaken opinion was a caution against viewing the world as computable. The latter was not heeding the caution.

Mother nature works in mysterious ways.

Pretty cool. I cant wait for bio-hardware to become the norm. The day of a laptop that has never ending storage that self heals (assuming there are no nasty laptop eating viruses/bacteria around) could be very handy.

Never-ending storage that self-heals... sounds a lot like cloud storage, innit? :-)

We're actually building a giant brain; the Internet /is/ a global collective consciousness. You should check out Serial Experiments Lain and Eden of the East :)

I would say that we had a global collective consciousness even before. Not to say that everything is not working much faster and beter.

We're actually building a giant brain; the Internet /is/ a global collective consciousness.

Well, we WERE building a global collective consciousness in the form of the internet, but the World Wide Web, and especially cat videos, put an end to it.

Or what if, instead of connecting biochemical to hardware, you connect hardware to your biochemical body. Add wifi and you're connected to everything, anytime you want.

Exactly - you just need to think "What's the square root of 459683" this queries the internet returns the result and the answer is instantly relayed to your mind. Just imagine you could think (open the garage door) and it would happen.

after some of the days i feel we all go though, it concerns me that people might know how many times i wanted to stab someone...

lets shelve this, we're already failing at online privacy...this wont help...unless i can think "this is a private thought" and it actually be private haha

I'd say everything is private by default and we actively have to make something public. Like sand boxing. Oh man, never mind.

Having to formulate the query is too primitive ;) When you need to know the result, the brain probably does a memory "lookup" subconsciously (if I see e.g. 3x3, I know the result without having to calculate it), so that could be routed through the digital system and the result returned as if it was a memory. You'd just "know".

Prepare to be assimilated. Resistance is futile! --

On a serious note, I love the technology like any other guy. But there will be this nagging issues about privacy. We would need to give up some parts of our lives, but who draws the line and where..

I am more concerned about the inevitable bugs, and failures that anything has. Except now it's brain connected to my brain.

But you can turn a hard drive off and stop having to feed it power, then come back to it months later to recover the data. You can't stop feeding biohardware, it would decay and decompose.

Sounds like the idea recycling program to me! haha

Do you want a laptop that remembers what you want or that _approximately_ remembers what you want ? Photos, Movies and songs may make good storage media for bio-hardware, but documents and other information critical stuff can not be replaced by biology.

It would be significantly more interesting if they were 6 _qubit_ bytes.

I was just about to comment upon the inclusion of Hameroff in the author list. Penrose and Hameroff are pushing a notion that the brain utilizes quantum coherence to perform computation inside these microtubules, resulting in human consciousness.

There is very little evidence for this conjecture, but Penrose and Hameroff are convinced. Make sure to keep Hameroff's bias in mind when reading this paper...

Regarding bias, I read somewhere that Numenta claims that the can built conciousness.. however regardless of that it's still quite interesting to see progress in Neural Networks.

They say: They found the key for a new learning algorithm that automatically finds patterns in streams of data and predicts what is likely to occur next.

more http://t.co/cVXRj0F6 and http://t.co/seWjurbW

It "is" possible that it's 6qubit bytes. I've read a lot regarding that, just follow the links http://en.wikipedia.org/wiki/Quantum_mind

So my brain uses base64 encoding? Didn't see that coming.

Reply on news.ycombinator.com