--

Neanderthal DNA Sequence Predictions

I had a very nice vist to my blog and email from Don Johansen who discovered Lucy in 1974 [not Simon]:

http://www.annonline.com/interviews/961204/biography.html

He was checking out my Neanderthal DNA Sequence predictions and notes on the relationship between autistic spectrum, brain volume and hypersurface area that I just wanted to brag about. The Neanderthal genome is being sequenced now...so we'll know soon if there is any merit to my predictions...

http://www.mindpixel.com/chris/2005/07/neanderthal-dna-sequence-predictions.html
http://www.mindpixel.com/chris/2005/07/neanderthal-theory-of-autistic.html
Permalink Chris McKinstry 
August 25th, 2005
Y'know, I was vaguely irritated a while back at the fact that ?off was becoming muppet's personal blog. I mean, if ?off is going to be anyone's personal blog, I thought, why -muppet-? God, why? What could be worse?

With this post, I officially apologize to muppet.

Cheers!
Permalink Snark 
August 25th, 2005
Oh come on. Would you rather read about sucking other people's blood? Bosses who won't play golf with their blod sucking underlings? And the blood sucker's diet? Or Neanderthal DNA?

Well, I guess the choice is yours.

But if you can tolerate Muppet, there is fuck all you can do about me.

I'm here.

I'm just like muppet except slimmer, smarter and sexier.

With a WAY BIGGER EGO.
Permalink Chris McKinstry 
August 25th, 2005
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!1
Permalink Aaron F Stanton 
August 25th, 2005
God damn mother fucking chimp sucking absolute dumbass loser fuckwad twit.

Jesus fucking christ go annoy somebody else for a change.
Permalink Aaron F Stanton 
August 25th, 2005
Now way.

I like it here.

Blame Muppet.
Permalink Chris McKinstry 
August 25th, 2005
I really so wish you'd just find yourself talking to yourself. A simple task for any moderator.
Permalink Aaron F Stanton 
August 25th, 2005
Yes, and the hyper-spatial trans-metamucil topographical mappings of interstital mind-space suggest that the forward-reverse, top-spin, left-chiral morphogenic field is slowing in its angular momentum somewhat. But it could be an inpecunious result of plebicite absorbtion in rubik-space. Either way, it's just hypothetical for now.

Wait, what the fuck am I saying?
Permalink John Haren 
August 25th, 2005
I kinda like it. It's like being in a Monty Python skit.

Think, dead 7 dimensional parrot mind ...
Permalink Mongo 
August 25th, 2005
Aaron. How many thalmocortical layers do you have? A million?
Permalink Chris McKinstry 
August 25th, 2005
Yeah, a Monty Python skit, but without the point. Or humor.
Permalink Aaron F Stanton 
August 25th, 2005
I couldn't give a rat's ass how many layers there are right now.

I bet you could really use a great algorithm for global optimization of functions. Excellent for high dimensional spaces. I wrote one. Beats the hell out of all others at time of publication.
Permalink Aaron F Stanton 
August 25th, 2005
Really. Like multi-dimential scaling?
Permalink Chris McKinstry 
August 25th, 2005
It scales extremely well with number of dimensions, far better than a genetic algorithm.
Permalink Aaron F Stanton 
August 25th, 2005
GAs are too broad a class to make that kind of cliam.

My question was more specific. How does your idea compare to MDS[1978]:
http://en.wikipedia.org/wiki/Multidimensional_scaling
Permalink Chris McKinstry 
August 25th, 2005
Chris,

Some advice on making the pain stop.

http://www.bmezine.com/news/people/A10101/trepan/
Permalink Michael B 
August 25th, 2005
http://www.pavis.org/essay/multidimensional_scaling.html
from that article is a good survey of various optimization methods. Deterministic methods get stuck far too often, and scaling of GA's is well established. Simulated annealing is most similar to the method I developed, but converges too slowly.

So yes, I did a full literature survey at the time. It's part and parcel of getting a doctorate, you know.
Permalink Aaron F Stanton 
August 25th, 2005
Thanks, I am way ahead of you:
http://www.mindpixel.com/chris/2005/06/ancient-brain-surgery-revived.html

Now someone just help me hold Muppet down.
Permalink Chris McKinstry 
August 25th, 2005
Hey Michael -

Thanks! I had forgotten that article. del.icio.us'd now.
Permalink Aaron F Stanton 
August 25th, 2005
Aaron. There is a specific technical question on the table.

Plan on answering?

I bet you didn't even know about MDS.
Permalink Chris McKinstry 
August 25th, 2005
>With a WAY BIGGER EGO.
With that last post, Chris' ego acquired its own force of gravity, breaking the seal on his Mr BlackHole(TM) waste disposal unit, sucking the rest of the his apartment and lab into nothingness.

Although the Sirius Cybernetics Corporation claimed their Mr BlackHole(TM) line of consumer black holes was avocado in color (and completely safe), Chris was unable to obtain satisfaction with his lawyer (or his dueling keyboards) as the SCC had been first against the wall when the revolution came. Curiously enough, when the New Improved People's SEC starting investigating SCC, they discovered that all the records had been dumped, Enron style, into one of their own black holes. The NIPSEC were never really able to determine if the black holes were indeed avocado, cerulean blue, goth black or stainless as they were somewhat invisible, like those pesky physicists claimed all proper black holes would be. Leaving the tinfoil hat parade to claim that NIPSEC were trying to cover something up.
"We don't know what they're hiding, but they're up to something!" claimed the anonymous spokesperson for the TFHP. "And where do you get off asking me for my name? Are *you* one of *them*?"
Permalink Peter 
August 25th, 2005
Oh man, here I am feeling like a jerk for suggesting that Chris McKinstry could find peace by trepanning himself, just like the guy in that Pi movie does...

(slaps forehead)

http://groups.google.com/group/sci.math/browse_thread/thread/c725660a8620cd43?hl=en&lr=&ie=UTF-8&rnum=206

Christopher McKinstry  Oct 21 1994, 3:38 am  show options
Newsgroups: sci.math
From: cmc...@access.mbnet.mb.ca (Christopher McKinstry) - Find messages by this author
Date: 21 Oct 1994 05:28:48 GMT
Local: Fri, Oct 21 1994 1:28 am
Subject: pi record
Reply to Author | Forward | Print | Individual Message | Show original | Report Abuse

-what's the record for calculating pi?
-is there someplace i can get an ascii file with atleast 1million places
of pi?
-has anyone scanned pi for messages from god? there must be some neat stuff
in there if you just look right...
Permalink Michael B 
August 25th, 2005
Waaaaaaahahahahahaah
Such fun ^^
Permalink Masiosare 
August 25th, 2005
Honestly, no, never heard of MDS. Given what wikipedia tells me, it is itself not an individual algorithm but a class of algorithms, your question is irrelevant. Studying the link I gave you reveals that I did compare my method to the bulk of methods described, and found it superior, which is why I answered as I did.
Permalink Aaron F Stanton 
August 25th, 2005
Unfortunately I don't have my thesis handy, or I'd post its list of references.
Permalink Aaron F Stanton 
August 25th, 2005
You're free to google for my published papers, though. They're on my advisor's website, neatly collected, and almost all references in my thesis are contained in those papers.
Permalink Aaron F Stanton 
August 25th, 2005
Nevermind, I'll save you the effort:

http://power1.chem.purdue.edu/kais/pivot.html
Permalink Aaron F Stanton 
August 25th, 2005
So. Again. Aaron. How did your program work?

And how it is better than GAs [which are an open class - new ones are being invented everyday - which is why I objected and still object to your boast].
Permalink Chris McKinstry 
August 25th, 2005
Feel free to object. Given that it was better at the time, and I haven't touched since I graduated, I should hope that someone else has improved on it by now.
Permalink Aaron F Stanton 
August 25th, 2005
Oh yeah...how did it work?

Been waiting for that.

Go fuck yourself.

Read the damn papers, asshole.
Permalink Aaron F Stanton 
August 25th, 2005
Hey Chris, no matter what everyone else says, it's cool you heard from the guy. Can you tell him something for me? Tell him I like his new show where he plays a lawyer, and I'm glad to see he's not wearing those white jackets over t-shirts any more.

And how's Philip Michael Thomas?

Philo
Permalink Philo 
August 25th, 2005
Hey, relax I am reading the papers. You posted while I was writing.

So, a couple of questions. What is to stop a GA from discovering your method? And if you know how to throw around random probes in a phase space why on earth are you giving me so much heat about my work that models the brain as a 7d phase space? My Mindpixel corpus is actually nearly 2 million random probes of human semantic phase space.
Permalink Chris McKinstry 
August 25th, 2005
You'd think a guy who's juggling all that stuff would notice a broken image link on his site. Or, if it really worked, it would TELL him there's a broken image on his site.

Philo
Permalink Philo 
August 25th, 2005
What broken image link?
Permalink Chris McKinstry 
August 25th, 2005
Ask your 7th-dimensional psychic hotline.

Philo
Permalink Philo 
August 25th, 2005
Ah. More brilliance. You're going to start licking other people's blood off your desk and start complaining no one will intite you to play a game you don't know how to play any second now.

Is your head starting to take on a felt-like texture? If so, I'd see a doctor immediately! And go on a diet.
Permalink Chris McKinstry 
August 25th, 2005
"I'm just like muppet except slimmer, smarter and sexier."

I'm sad to say that the two of you could, at a quick glance, have been seperated at birth. For visual confirmation, I've hosted his disturbing video here:

http://dashslot.co.uk/media/mupparody_recode.mpg

I just wish I was here earlier, but someone beat me to the obvious Miami Vice jokes. In the spirit of providing information, though, Philip Micheal Thomas was pimping some kind of lame psychic hotline a 5-odd years ago. Where he is now I couldn't say, but "near the bottom of the pile" is probably a reasonable estimate...
Permalink Mat Hall 
August 25th, 2005
GA's don't create new methods unless you're doing some sort of dynamically-self-recoding optimization algorithm. They don't search algorithm space, they search phase space.

Which, actually, I did experiment with a little bit, in adjusting parameters for my method.

It's not your 7 dimensional space I object to, though I do think that's pretty absurd.

My point is that you're exactly as special as the next Fight Club snowflake in the room. I've met Dudley Herschbach, Ralph Merkle, and Eric Drexler. I have 3 published papers. I have created something that if people used it rather than older slower tech, their stuff would be vastly more efficient. (Sounds familiar? Like you a bit?) And yet, I'm no more amazing than anyone else here.

You're in a room full of really smart people. Many have done some really cool stuff. So give the messiah complex a rest.
Permalink Aaron F Stanton 
August 25th, 2005
Hey, Aaron that's cool. And it is wonderful I provoked you to talk about it. I wish more people would just talk about what they have done.

As for GAs...there is no rule to say you can't nest them - GA's looking for GA's...GA's in general, can do anything that can be done.
Permalink Chris McKinstry 
August 25th, 2005
Of course they can do a lot, but so can other optimization methods.

Mine, of course, would do it better.

;)
Permalink Aaron F Stanton 
August 25th, 2005
"And yet, I'm no more amazing than anyone else here."

damn. you are *definitely* more amazing than me. so cool, Ive always wanted to study that stuff.

bastard.
Permalink Jesus H Christ 
August 25th, 2005
Here's why I went to grad school: Molecular nanotechnology. I heard of it around '90 and decided it was so incredibly cool I just had to try to get into it. I got my shit together, worked about 66 hours a week for 9 months straight to pay bills enough to go back and finish my Bachelor's. Then I applied to grad school. Purdue said "We'd like to take you, but we're out of money this year, so please reactivate your application next year." So I did. Drove there to visit a friend once, personally stopped by the chemistry department to check on the status of my application, and they took it out of a filing cabinet and put it on top of a pile. Next thing I know, I'm in grad school, summer of '94. Then it's just stay in classes and stubborn it out until I decide I am done, write everything up and leave with my PhD.

And then fail to find a job.

That's the fun part. Oh, there's stuff in the middle, like how partway through the program the head of the department informed me I was getting a degree in chemistry, not nanotechnology, where did I get the idea I could do that - despite the fact that he was the very person who had told me I could when I entered two years before. Or how I got blown off at a (nanotech) conference when I told people I was a grad student interested in studying nanotech, despite the fact that grad students are kind of necessary for a field to grow. Or how people couldn't handle the idea of machine phase chemistry. Or how one professor laughed at it, and now has published a book on it, pulling a complete 180.

Bitter? Me?

Never.
Permalink Aaron F Stanton 
August 25th, 2005
wow, what a whinging bastard :)

...nice to know that if I had chosen a different life path I would probably still be equally bitter about the way things turned out...
Permalink Jesus H Christ 
August 25th, 2005
That was great Aaron. Really. Your method may in fact be better than every known GA in some domain.

But say a GA managed to discover say, a tunable fractal version of your random probe? And then suppose that the attractors in the phase space you are trying to map are also fractal?

I say I just discover it?
Permalink Chris McKinstry 
August 25th, 2005
Yeah, I usually don't talk about it. I guess I decided that the degree was where I'd stop, and then went no further. I probably should have stubborned it through to the point of getting a job in the field, but that's the past.

Time to move along to the next thing, eh?
Permalink Aaron F Stanton 
August 25th, 2005
Chris -

A GA might discover that, but you seem to have missed what I said. If you can construct a method of searching algorithm space with a GA, you can also search that space with my method.

I actually *did* do this to a limited extent. Because the earliest implementations of it required some arbitrary parameters to be set, rather than just guess, I nested it inside itself and ran the outer optimization loop until the parameter in question stabilized.

If a GA can discover my method and improve on it, great, so be it. But my method could do the same thing. A GA varies elements by mutation and point swapping a given vector within a population, among other operators. My method is similar. It is in some ways a mix of GA and simulated annealing.

I understand how useful fractals are. Non-integer dimensions used to be extremely interesting to me, particularly in conjuction with nanotechnology, but that's beside the point.
Permalink Aaron F Stanton 
August 25th, 2005
> there is fuck all you can do about me

You know, some people might take that as a challenge.
Permalink  
August 26th, 2005
> I got blown off at a (nanotech) conference

That stripper again?
Permalink  
August 26th, 2005
Funny, but no.
Permalink Aaron F Stanton 
August 26th, 2005
"I understand how useful fractals are."

oooo.... shiny....
Permalink Philo 
August 26th, 2005
Yeah, fractals tend to be nice shiny objects, but from time to time they come in handy.
Permalink Aaron F Stanton 
August 26th, 2005
Aaron, Sabre Kais did all the work and you fucking know it!
Permalink stop the lies for once!!! 
August 26th, 2005
Hey Aaron...I am super interested in this topic. I will come back here and pick it up as soon as I can. I have a pile of work today.
Permalink Chris McKinstry 
August 26th, 2005
Lol.

Sabre didn't look at my source code even once, not even in my thesis. Now, Pablo Serra did a respectable portion of the work on two of those three papers, and he's simply amazing, but my advisor? No, he didn't really do much on those.

Sabre's blindingly intelligent. I wish I had been involved with some of the work they did mapping quantum mechanics to thermodynamics, but I was otherwise occupied at the time.

Incidentally, in most of the academic world that I've seen, last name on a paper generally means the person running the research group, or if more than one group is involved, it's the most senior person, usually by tenure. I believe we followed that convention in our papers.
Permalink Aaron F Stanton 
August 26th, 2005
Chris -

I knew you would be interested, which is why I baited you with it.

As long as you remain not so much of a pompous ass, we'll get along just fine.
Permalink Aaron F Stanton 
August 26th, 2005
fap fap fap
Permalink muppet 
August 26th, 2005
muppet, darling, I know you're excited to the point of ejaculation by all this, but could you save a little for me?
Permalink Aaron F Stanton 
August 26th, 2005
So, Aaron, when this McKinstry guy can go to the library, grab a book on [really complicated subject], read it, do some Google, proclaim himself an expert, and troll-market himself by babbling the shit out of the subject so that anyone without a basic introduction thinks he must be a genius and respects him as some kind of scientist, it must make your fuckin' blood boil because it looks like you did actual real work whereas he just read half of the wikipedia article on nanotechnology.

Need a hug?
Permalink Michael B 
August 26th, 2005
*snicker*

You're not far off the mark, actually. Snake oil salesmen piss me off.

No hug required. Thanks, though.
Permalink Aaron F Stanton 
August 26th, 2005
Ok. New week.

Back to phase space and fractal sampling thereof...

And my question is, what do you find silly about modeling immediate memory as a fratal pattern of activation on the surface of a seven-sphere?
Permalink Chris McKinstry 
August 29th, 2005
Michael B.

There aren´t many libraries with the kind of books I need down here in Chile atually. I had to buy most everything:

http://www.amazon.com/exec/obidos/ppl/A3PK2LTN51JAMF/ref=cm_pdp_more_rp_shared/103-7857358-1352625

Then it took five years of work by myself and more than fifty thousand people to get the data set I am using in my experiment now.
Permalink Chris McKinstry 
August 29th, 2005

This topic was orginally posted to the off-topic forum of the
Joel on Software discussion board.

Other topics: August, 2005 Other topics: August, 2005 Recent topics Recent topics