8.3.10

evil plan results:

Bond loses. in other news, the implant has healed properly, to my surprise, and is now safely lodged in my left pinky and functioning at what i guess to be about half of its nerve-triggering capacity, based on what the others do.

i'll also be writing another article for H+ as previously hoped, on the ethics, procedures and legality of self-performed surgery (minor). thoughts?

L

30 comments:

The white crow said...

Include demonstration drawings?
Talk about what you'd do if you had the opportunity of using OR surgery and any materials (with current technology). Always interesting to see and know what could be done differently, given the right conditions.

Anonymous said...

Dude, your fucking awesome, keep up the good work.

Onesan sama.

QQBistro said...

I understand what you're doing and it makes me very happy.

Lepht said...

Crow - it's gonna be a practical article rather than a theoretical one, since H+ is kinda already full of people speculating on what they might do but can't. the drawings are a good idea.

QQBistro - thanks. it makes me happy too.

L

Ian said...

If it's going to be practical, maybe you should slip something about how you're going to implant that NorthPaw gadget into your leg (or have you done that already?).

~Ian

Lepht said...

haven't gotten to it yet; still trying to find high-quality shit for output. point, though, i'll need to put that in. cheers

L

MooNWalker said...

What high quality shit for output you are talking about? Isn't there already interface based on vibrating motors?

Lepht said...

electrodes; can't seem to find any good ones, since i wanted neuro-quality. might have to downgrade. the kinetic motors are kinda... clunky.

L

MooNWalker said...

I'm afraid just randomly putting in electrodes may not do any good, would need to find some nerve endings to connect them to... Do you have anyone you know from medical/neuro etc. who can say more on this topic? I don't know if you are familiar with works of Kevin Warwick, but he performed experiments on integrating electrodes with nerves. Plus he is nearby, in England, so you may want to talk to him about collaboration. Check http://www.kevinwarwick.com.

Lepht said...

yeah, i was trying to find the nerves... you're right, it's gonna be tricky. my source for neuro-type information was unfortunately the same friend who tried to have me sectioned, and has now stopped talking to me. i'm still pretty fucked up over that.

ah, Warwick. you're aware he's gone nuts, right?

L

MooNWalker said...

What do you mean? I'm not very good at following the news about what happens to people.

Lepht said...

he's gone nuts in that he thinks he's a cyborg for a start, but also that he is going to save the world from evil robots and we all need to become cyborgs like him to fight the robot army. i am not even shitting you.

still, his electrode work is valid. might have to look that shit up.

L

Ian said...

i personally would love to become a cyborg (though, technically, anybody with any gizmos implanted in them fits the definition), but that's probably not going to happen in the sense that most people think about within our lifetime, despite what Ray Kurzweil thinks.

~Ian

MooNWalker said...

There is no point in debating what will and what will not happen. If you want it, go and get it - work in that direction. If not, than just leave it. No one needs that pessimistic attitude, it doesn't help to get the job done.

But Ian is right about cyborgs - technically Warwick is a cyborg, though not the first one as he claims. There are plenty of people who got their cochlear implants long time before Warwick even started his work. About robot army - well, there are millions of people that believe that evolution is a hoax, and honestly, I think those people are much more dangerous.

Unqualified to speak said...

Maybe electroacupuncture or EMS would be a good starting point for an interface; muscle stimulation, meaning you don't need to worry about getting a neurology degree. As long as you would feel the muscle twitch, and it doesn't fuck with limb use or something...

Ian said...

@MooNWalker-

i agree that we should work in that direction, and i very much plan to, but it's just that Ray Comfort keeps getting everyone's hopes way up-he actually thinks that nanotechnology will cure ageing and end every affliction known to humanity by the 2020s, which, while we should work towards, probably isn't gonna happen.

also, that would be an interesting question: what is more dangerous to humanity, a robot army bent on wiping out the entire human race, or these pseudoscientists which basically threaten to undermine the very foundations of society?

~Ian

Ian said...

oops, srry meant Ray Kurzweil, not banana-man. he's one of the ones who fit into that pseudoscientist category.

MooNWalker said...

"threaten to undermine the very foundations of society"
And what are those foundations, I'm kinda curious now?

Lepht said...

Unqualified - yeah, that's more or less what i was thinking. the problem seems to lie in finding any supplies at all; i need more research, tbh.

MooNWalker, Ian - agreed, pessimism != useful; i think all of us are fairly optimistic about technological progress, though. hell, we're involved in it.

L

ps. don't even get me fucking started on that dreamer quack Kurzweil...

Ian said...

yeah, i am pretty optimistic in terms of the fact that I think that the developments of technology that even people like Kurzweil are writing about, i just disagree on when it will happen. although, he did get one date right--the approximate date of the singularity. if you extrapolate Moore's Law out to the 2050s (which you really can't do, since said law will collapse in ten years. this however will likely be solved due to some other computer such as quantum, optical, peptide, DNA, etc. which we are making great improvements on), you will get a smarter-than-human intelligence, which is not too far off from Kurzweil's date of 2045.

Unqualified to speak said...

@Lepht: a cheap TENS unit (by the sound of it, EMS as well) any use to you?
@Ian: good ol' silicon might well get us a good chunk of the way there.

Lepht said...

Unqualified - maybe, although i was looking for subdermal electrodes, TENS doesn't seem to give much of a buzz and the trodes wear out really fucking quickly.

Ian - but you can't extrapolate Moore's out that far in any case; that's why he's a quack...

Unqualified to speak said...

Yeah; I was thinking as a cheap "would it work?" first iteration. Don't know nuffin' about subdermal electrodes, I'm afraid - would the transdermal needle ones for EMGs n' stuff be any good?
Or just get a set of transdermal piercings and a soldering iron... (",)

Unqualified to speak said...

Damnit. Sorry about that; Blogger and NoScript don't like each other very much...

Ian said...

Unqualified-i agree, silicon has quite a ways left to go in the ten years it has left, but, as Lepht alluded, in ten years we will no longer be able to increase the power of silicon computing.

the reason that silicon computing power doubles every 18 months is because we can etch smaller and smaller chips onto the silicon wafers. right now your pentium chip (or whatever chip you happen to be using) contains a layer about twenty atoms across. according to moore's law, in 2020 that chip will have a layer five atoms across. at that small a scale, we have to abandon classical physics and use quantum physics instead, where we must use the heisenberg uncertainty principle, so we don't know where the electron is-it could be in the circuit, outside, both, or neither-in other words, we have a short circuit. people have said that moores law is "the victory of bits over atoms," but in ten years, atoms may have the last laugh, especially if quantum computing replaces silicon computing.

Ian said...

Lepht-i agree, we can't extrapolate moore's law out nearly that far, but we are making steady advantages on new types of computing, such as:

DNA: a turing computer using DNA as its input/output tape.
Advantages: it provides a very simple way of solving the "traveling salesman" problem: given several cities, what is the shortest route to reach every house in those cities?
Disadvantages: unlike copying on silicon computers, dna does not copy itself perfectly; we see the effects of this all the time when we age. since half of computing is based on copying information, this is a serious setback.

Peptide: computing on proteins and amino acids
Adv: rather than using base 4 math, since there are 4 dna bases, peptides can use base 20, since there are 20 amino acids, greatly increasing the potential complexity of programs it can handle.
Disadv: as great as they are at catalyzing reactions, proteins are very poor at storing information.

optical: silicon computing using photons instead of electrons.
Adv: as opposed to current silicon computers, which are composed of flat "wafers" stacked on top of each other, optical computers have a 3d structure, greatly boosting complexity.
Disadv: they are subject to the same limit modern computers are: you can't miniaturize them too much without them shortcircuiting. also, it is nearly impossible for photons to react with each other (which is a good thing, since vision would not work if they did)

chemical: basically a dish of chemicals
adv: it is much cheaper to produce than any other computer (all you need is a bunch of reactive chemicals and a dish)
disadv: eventually, any chemical reaction will reach equilibrium, at which point you can't get any processing out of it. these computers would need either a constant input of chemicals or a robotic hand that, as New Scientist puts it, "literally 'stirs' the creative juices."

quantum (my personal favorite): computations made on individual atoms
adv: instead of using conventional bits, which only take the values 0 and 1, these computers use "qubits" which can take any value between 0 and 1, much like fuzzy or probabilistic logic. this is extremely promising. for example, it is very hard to factor numbers on conventional computers, which makes todays codes nearly impossible to break, but it would be childs play for a quantum computer to use shor's algorithm to break these codes with complexity only equal to O(sqrt(n)) in the worst case, where n is the length of the number being factored. it only takes less than 1 qubyte to perform the calculation "3x5=15."
Disadv: these computers require that all the atoms are coherent; that is, vibrating in unison, which is extremely hard to do for large numbers of atoms. That 3x5 calculation is the world record for a quantum computation.

All things considered, with the imminent collapse of moores law, and the progress made in other forms of computing, i'd say that the singularity will happen around 2100. not in our lifetimes, but much closer than most people think.

well, that was rather lengthly, but i felt like showing off a little bit, and i think that it answers the points made by both users rather well.

~Ian

MooNWalker said...

Ian, I think you're a little bit off in your reasoning and way too obsessed with performance. First, singularity is not all about pure computer performance, it's about means to perform the development more effectively, which only partially depends on raw processing power and much more on interfaces, ways of interacting, communications.

Second, about Moore's "law" - let me remind that the original formulation of Moore's law was stating that number of components per IC (integrated circuit) at which the price per component is the lowest will be doubling every year, later he changed it to two years. It says nothing about performance, it says nothing about the size of components, it talks about economical suitability.
But lets forget about that and assume that the size of die stays the same. Then for the number of elements to double every two years size of elements needs to decrease twice per four years. There is already discovered semiconductor of the size less then 1x4nm, that makes 5nm process quite feasible, correct me if I'm wrong. That gives us three decades instead of just one, that's plenty of time to get the shit polished shiny.

Ian said...

M00nwalker--why shouldn't i be obsessed with performance? the singularity is supposed to be the moment at which we create a superhuman intelligence, which will be able to create more intelligent versions of themselves, ad nauseum, resulting in a technology explosion (this is at least how Kurzweil defines it in both his The Singularity is near and The Age of Spiritual Machines). so, i should be focused on performance, because that's what is going to kick off the initial event.

second, Moore's law describes a long-term trend in the history of computing hardware, in which the number of transistors that can be placed inexpensively on an integrated circuit has doubled approximately every two years. this translates approximately to a doubling of computer power every eighteen months--power is not directly proportional to number of transistors. a silicon atom, by the way, is not 1 nm across; it is 111 pm. again, we probably will not be able to reduce the scale to 5 atoms, which is 555 pm, not 5 nm, without shortcircuiting.

like i said, though, in the decade we have left, there is a lot of room for increasing power, so silicon will probably get us a good chunk of the way there.

~Ian

MooNWalker said...

First, let me correct myself - 1x4nm is not semiconductor but superconductor. Second, problem is actually opposite to short-circuiting, - problem is with high resistance of standard metal connectors between elements on die, they can easily overheat and "melt". Molecular superconductors is the future solution to this problem and their size is the limit for the process, not the size of the silicon atom.

About pure performance - just research history of processors development, especially about "pure performance" of the Pentium 4. And also think about reasons behind trends of having multiple cores on one die.

MooNWalker said...

BTW, Ian, I would appreciate if you don't put zeros in my nickname. It's moonwalker, not m00nwalker.

Post a Comment

[pls no ask about the vodka. debate is always welcome. remember, Tramadol fucks you up]