Engineered Organisms

[DakAttack]

Re: Engineered Organisms

Because it's becoming easier and easier every day for those fantasies to start translating into realities.

We would have to design AIs more complex than our own minds for any of this to happen. How do you program an AI to want more? To know there's better out there? To innovate? To formulate novel ideas? None of this is possible.

A program can't do anything it wasn't written to do. That's the bottom line.

[Taskmage]

Re: Engineered Organisms

How do you program an AI to want more? To know there's better out there? To innovate? To formulate novel ideas? None of this is possible.

The first isn't even necessary because "want" doesn't have to be part of the equation. Just program it to keep doing. Likewise for the second. The latter two are harder but by any means impossible. Everything that exists does so within a framework of discrete systems. The Eureka AI deduced Newton's 2nd Law just by watching a double pendulum system and analyzing the data. Humans took thousands of years to "invent" that concept and this still quite limited AI did it in two hours. From there the combining of newly discovered concepts into working designs is a matter of integrating them into genetic algorithms that humans already use computers to simulate as part of our own design processes. The groundwork for much of this is already in place.

---------- Post added at 04:52 PM ---------- Previous post was at 04:43 PM ----------

Say you write a simple calculator program. You tell it how basic arithmetic works and then you ask it questions. It can easily give you answers you don't know and couldn't have reasonably worked out for yourself, like the product of two 20-digit numbers. Likewise a program could be made that had a basic framework of material science that could give us answers to questions about microprocessor design that we didn't know the answers to before making the program. The program wouldn't be doing anything it wasn't programmed to do, but it would be creating innovation. Hell, this is already how innovation works. The only difference now is that humans are the ones identifying gaps in our knowledge and using computers to fill them. When a computer can have a knowledge base, identify unresolved questions and seek the answers itself, then it can innovate on its own.

--------- Edit --------

More on Eureka: http://www.guardian.co.uk/science/20...ntelligence-ai
More on genetic algorithms: http://brainz.org/15-real-world-appl...ic-algorithms/

[yavimaya]

Re: Engineered Organisms

Totally read the title of the thread wrong.

I am glad I am not the only one.

[DakAttack]

Re: Engineered Organisms

The first isn't even necessary because "want" doesn't have to be part of the equation. Just program it to keep doing. Likewise for the second. The latter two are harder but by any means impossible. Everything that exists does so within a framework of discrete systems. The Eureka AI deduced Newton's 2nd Law just by watching a double pendulum system and analyzing the data. Humans took thousands of years to "invent" that concept and this still quite limited AI did it in two hours. From there the combining of newly discovered concepts into working designs is a matter of integrating them into genetic algorithms that humans already use computers to simulate as part of our own design processes. The groundwork for much of this is already in place.

So somebody programmed a computer to do something and it did it? Something cavemen couldn't do with their cavecomputers? I better go back in time and tell them to step their game up.

Say you write a simple calculator program. You tell it how basic arithmetic works and then you ask it questions. It can easily give you answers you don't know and couldn't have reasonably worked out for yourself, like the product of two 20-digit numbers. Likewise a program could be made that had a basic framework of material science that could give us answers to questions about microprocessor design that we didn't know the answers to before making the program. The program wouldn't be doing anything it wasn't programmed to do, but it would be creating innovation. Hell, this is already how innovation works. The only difference now is that humans are the ones identifying gaps in our knowledge and using computers to fill them. When a computer can have a knowledge base, identify unresolved questions and seek the answers itself, then it can innovate on its own.

So we program the programs to use themselves. How would we give them access to their own data to allow them to rewrite themselves? How would they even know how that language works?

[Taskmage]

Re: Engineered Organisms

I added links to my previous post. One explains more about Eureka. It was given some data sets collected from real life and asked for any interesting truths it could deduce and it spat out some laws of motion and conservation of momentum. It wasn't programmed to find Newton's laws.

So we program the programs to use themselves. How would we give them access to their own data to allow them to rewrite themselves? How would they even know how that language works?

It's as easy to give a program access to its own code as it is to give it access to anything else. The second we would have to program.

[DieselBoy09]

Re: Engineered Organisms

What's so wrong with all of this? In the first part when they talked about the engineered animals, most of those are sterile, if not all. It is nature's way of kind of keeping thinks like that in check to a degree.
As for the bio-luminescence? really? I've done that with yeast and bacterial colonies in my beginner bio labs. Its not a difficult process, its just making the cell competent to take up a plasmid, which is often some what of a crap shoot. Also, they don't literally glow that color when the lights go off, you need to shine UV on them.
As for genetically altered foods.... you don't take up genetics from food, so.... not a big deal so long as it doesn't influence the genes in wild viruses which can then potentially infect us. Now as for cloning, cloning in general sucks. When you clone something you take a cell from that generation with all of its flaws and mutations and bases an animal off it. It has a shitty life span and ages prematurely. It is simply the nature of the cells used. Unless one is cloned from the first generation cells, the clone will not be worth it.
As for knockout genetics, they're awesome. No harmful effects really. Saving endangered species with clones, sure, awesome, so long as you're not cloning clones. As for the mixed genetic DNA, mitochondrial and cellular, that's not a concern at all.
As for the machines controlling the bugs.... its not really control when you're just stimulating nervous cells to influence behavior.... its kind of strongly suggesting, but they're not completely controlled. As for the prosthetic arm monkey, the hopes of such technology is to create artificial arms for injured soldiers and such, also to understand how the brain controls the body, its great and a necessary thing to do. Don't be scared to have this abstract idea of a magical existence thrown away and replaced with a functioning model of consciousness, thought and emotion.
Now finally on to the lamprey brain.... really? Lower organisms have much simpler nervous systems and brains, and so therefore far easier to achieve this feat with than a higher animal such as a human. And no, lamprey, like most lower organisms probably don't have thought, they lack a prefrontal cortex which is the seat of higher cognition. As for the ear, its amazing but not really an ear. Cells grow around scaffolding, in fact, they hope in the future to create working organs this way... however, a scaffolding provides shape, not necessarily proper functioning, and thus why that rat created a ear shaped appendage, rather than a working ear. And finally the artificial organism..... all of your cells run automatically based on the laws of chemistry and physics.. you exist simply because of all these things, rather than having some sense of control it is random but predictable, it runs on its own. The more you begin to understand about cells, chemistry, biochemistry, etc... it is hard to see what we view as life and really attribute a uniqueness to it. Viruses for example are not considered alive, but reproduce by taking over a host's cellular mechanism, its simply a chemical cascade which provides a simple example to the basis of life, despite not "being alive". When you're at home, take crisqo, and poor some in water... this will create a single lipid layer miscele, which is very similar to a cellular membrane due the the hydrophobic nature of lipids... congrats, you artificially created something similar to a cellular membrane! If you have a mixture of the components of a cellular membrane and poor them in water, likewise it will all arrange as a cellular membrane.

I do not care to take the time to write this all out eloquently, I have other things I need to do. However, do not let these technological "ethnicists" or futurists scare you with the nature of scientific advancement in universities or other institutions. While it is true that science is making amazing leaps and bounds, we are still decades if not centuries away from the fears of science fiction. As an example let me discuss the use of viruses to cure Severe combined immunodeficiency syndrome. In France several years ago, scientists genetically engineered viruses (I believe it was the herpes or adenovirus family of viruses) to infect patients with SCIDS, which would then leave part of their DNA upon the host (which viruses do), allowing for the creation of whatever cell or cellular product which was needed in order to develop a working immune system. When used, it worked. It completely cured SCIDS in the patients who were given the viral therapy. However, when running disease statistics on the patients and their immediate family, a form of cancer was found to occur in larger than natural percentages within this population, and thus the project was terminated for fears that the patients were in some way shedding viral particles which were then infecting family members and in some way inducing these cancerous side effects, despite there being not logic or evidence to back up such a hypothesis. Now the rate of cancer development in these groups was not outlandish, just slightly over the nation norms, but still enough to facilitate further investigation of the situation. Nothing has been found to this day about it, and in all likely hood, the cancers were not caused by the virus somehow transmitting to the family members and somehow, inducing cancer. And really, the project, what they did, was amazing, but even then, not all that advanced, and even though it did do a leap and bound with regard to medical treatment, levels of uncertainty further hindered the treatment and will till we understand more about the use of modified viruses to create cures. Moral of the story, its really not all that out there or something to be feared. Go to your local college and read up on some of the biology text books.

[Kailea]

Re: Engineered Organisms

my..head hurts....... *grabs his helmet* I am going to go fight some random doctor with a sweet stash, holophone me when my buster is ready.

[Firewind]

Re: Engineered Organisms

Totally read the title of the thread wrong.

Glad I'm not the only one.

[Feba]

Re: Engineered Organisms

The second we would have to program.

This has been done before, iirc. I've read of evolutionary programs which modify their own structure, see what does and doesn't work, the most effective adaptations survive, the weaker ones die off, basically exactly how it works with organics but on again a hugely faster timescale.

Computers don't yet have the ability to replace human coders. It won't be long before they do. Look at machine translation-- people used to think it would never work. Now, take a look at google translate. It's still far from perfect, but it's improved by leaps and bounds from what we were looking at on Babelfish ten years ago. Translation between European languages is to the point where it's actually quite readable. I've been to translator conventions where the entire focus of the day's events was "oh shit guys, look at this. People don't need us anymore, they just put what they need to read in google translate and get what they want.". They talked about how the machines have improved so much in recent years (google translate learns from comparing human translations of text, such as the Bible or UN reports, to see the relationship in how information is rendered, rather than interpreting grammar literally all the time), to the point where their entire field is going to change from that of "the ability to render text from a source language into a target language" to "confirming that the machine translation is accurate and readable". Prose will probably stay human translated for a long time, but save that computers are taking off there.

[DakAttack]

Re: Engineered Organisms

So when computers are able to modify their own code, in what ways would they be allowed to? Would we have to program the different ways or would we give them a database and let them go crazy? If we did give them a database, how would they know what they wanted to use, or why?

[Malacite]

Re: Engineered Organisms

Computers don't yet have the ability to replace human coders. It won't be long before they do. Look at machine translation-- people used to think it would never work. Now, take a look at google translate. It's still far from perfect, but it's improved by leaps and bounds from what we were looking at on Babelfish ten years ago. Translation between European languages is to the point where it's actually quite readable.

This is precisely what I'm waiting for. Fuck Cell phones, I'll wait until we have full-blown communicators.

[Aeni]

Re: Engineered Organisms

I think many of you trivialize the human brain. Ask the people behind Watson at IBM and they'll be the first to tell you that even duplicating the human brain is a feat that won't be matched for a very, very long time, never mind surpassing it. If you even want to approach the memory capacity of just 1% of the human brain, all the hard drives in the world wouldn't come close.

I think I remember reading about an article in the Scientific American which stated that the capacity would be around 2.5 petabytes, but then I think there was other evidence to suggest that might not even be close to the true potential because it made some very misguided assumptions about what data is and how do you measure that data. For all we know, the kind of memory we hold would be equivalent to an ultra high resolution image that may require 100gb to store each frame, since each frame could potentially hold additional sources of information that we are not consciously aware of, things like smell, touch, taste, etc.

[Kailea]

Re: Engineered Organisms

that is why before they can copy the brain, they have to learn how to interface with a real one, imagine being able to use your brain to store data, that you enter manually. Being able to download a language pack into your brain and then being able to speak, write, and read that language.

[Aeni]

Re: Engineered Organisms

that is why before they can copy the brain, they have to learn how to interface with a real one, imagine being able to use your brain to store data, that you enter manually. Being able to download a language pack into your brain and then being able to speak, write, and read that language.

I don't think it's possible to do that because as simple as it may sound, no one fully understands the complex chemicals and how they interact with organic matter. Therefore traditional approaches to engineering won't work especially since inorganic matter has some other issues, including unwanted reactions.

The better approach that I know they're working on at MIT and other labs is to clone the brain and develop organic interfaces vis a vis virii and bacterium. Nanotechnology will hold the key to interface artificially manipulated organic matter with inorganic matter that may be able to (indirectly) connect with the human brain ... but I contend that the human brain is far more superior and efficient at what it does. Putting aside the Matrix for a moment, I doubt any "short cuts" could be devised to input information directly into the brain. It would be far easier to create a "host body" which to "transport" yourself into that already has all of this stuff hard wired, i.e., Ghost in the Shell.

[Feba]

Re: Engineered Organisms

Being able to download a language pack into your brain and then being able to speak, write, and read that language.

language does not work that way. Go read up a bit on linguistics; the actual field is almost entirely about looking at language, and going "gee, now why do things work that way? BECAUSE BRAINS." There's a lot of complex connections and stuff going on; for a quick overview of what I'm talking about, look at wikipedia articles on aphasia. I'd imagine pretty much anything you learn to do would be the same way. Doing something like teaching someone a language would require something closer to remodeling the brain than just injecting new content. This is also related to the cutoff period for acquiring new languages natively.