Thoughts on Singularity?
http://www.time.com/time/health/article/0,8599,20…
What do you monkeys think about singularity?
More importantly, where is John Connor?
Trending Content
| +95 | Are you “less ambitious” for having long term goals outside of NYC | 24 | 4h | |
| +37 | Interviews Are So Fake | 24 | 5h | |
| +33 | 2024 UK Election - Tories finished? | 20 | 15h | |
| +29 | Being Christian in investment banking | 14 | 10h | |
| +26 | Is my boss gaslighting me? | 3 | 6d | |
| +25 | Non-Competes Banned | 17 | 3m | |
| +22 | Ideas for things to do with a free semester before starting ft? | 9 | 1d | |
| +20 | Sabotaging Peers to Get Ahead? | 34 | 3d | |
| +18 | Best NYC neighborhood for single 30M | 12 | 1d | |
| +17 | Moelis has the cutest Analysts? | 4 | 6h |
Career Resources
You're a bit late to the party bud - we just did this one. http://www.wallstreetoasis.com/blog/bonus-bananas-feb-11-2011
Sorry, I'm an idiot. Search function is key. I just read that one article and it completely turned my understanding of modern day economics upside down.
I saw it there too, but I'll comment:
In a way, the internet viewed as a neural net has long ago surpassed humanity and the collective interactions of the web are very quickly overtaking all other traditional means of information transfer...just look at Egypt: for this brief moment in history, who is really in control? The google guy? The people? US? At this point: no one. Only the information that reaches people. On the individual node level, even the crappiest 286 handily defeats any person in linear processing: but all they are doing is crunching a series of 0s and 1s. That's it. Computers will eventually catch up to us in parallel processing - the type that recognizes patterns and assocations, and allows for actual UNDERSTANDING. This is the root of sentience. This technology has not even reached a fledgling stage, so it seems unlikely we will see it in our lifetimes.
With web and/or computer based sentience, there is currently no central decision making center, no means of comparing this information against the physical reality outside of the electrical signals that are its thoughts, and no basis of comparison for its self identity. Computer sentience would be pure conciousness. Transferring our brain to such a system, while perhaps centuries away from being real, would be radically different than anything we conceptualize in todays terms: would I be able to download another person's thoughts and emotions? Would my online brain be able to interface with physical reality? Could I program it to give me a continuous orgasm, sugar rush, or mystical state? Could I merge completely with another conciousness or absorbe facets of another personality? Can I listen to ten songs at once? Will I be hearing, or simply downloading???? The capabilities of such a system are limitied only by our technological and ethical advances. By ethics I simply mean what we decide to create.....for whatever the reason.
Also, we project a lot of the way WE think as humans onto what may become computer sentience: if the intelligence is web based, it will need to keep the current infrastructure going, and killing us off would be against its own self interest, at least in the short term. Computers would not have the same range of emotions, survival instincts, physical senses, or desires as humans at all, so it seems that we are projecting our own likeness onto them. We should probably be careful of what we create, and ask ourselves not just IF we can, but also SHOULD we. Movies like T2 are great entertainment, but also a good warning of what NOT to build. What is also to say that such a creation wouldn't measure all existing medical information against itself and cure every disease in a few minutes? Or create 50 dimensional math? Or write a ten thousand hour long song? Or bring the entire global forex system into perpetual balance within a fraction of a second? Such processing power combined with understanding would be almost infinitely capable.
I come in peace. Take me to your bar.
1st Computers will only be able to do what some one programs them to do.
2nd Computers can't "learn" they run off of a program and store data, process it by instructions and spit it back. That is all.
3rd " it. Computers will eventually catch up to us in parallel processing - the type that recognizes patterns and assocations, and allows for actual UNDERSTANDING."
Do you mean programs that run them? You can have the Ne Plus Ultra of computers, but if it is running no OS you can't do shit with it. To create a program that "learns" is not possible. It would have to be able to write its own code and that just isn't possible, sorry.
In conclusion,
Personally, I see this stuff as so far down the line, perhaps centuries or thousans of years, that it's fun to think about for a little, but there's not much to be gained if you're not writing science fiction.
fun to think about if you're stuck up all night, yes?
Computers are currently only capable of doing what programmers tell them to do. What happens when artificial intelligence surpasses human intelligence? What happens when computers start programming themselves? What if computers become so much more intelligent than us, that they begin to developing their own programs at an exponential rate that makes today's version of Moore's law look slow, in comparison? What if cyborgs start developing hardware? What happens if we get attacked by cyborgs, when we try to unplug the computer?
Most people in the scientific community believe artificial intelligence is not as far fetched as many currently believe it to be.
I've got no choice, but to agree with Uncle Eddie on this one.
I'll be using that pic for some other random stuff, thanks.
np
I think the advent of quantum computing will allow many things that classical computing does not. I think once we build a quantum computer with a few billion qubits and really understand the power of quantum information, all of the things in the article will be possible.
Actually, just optical computing would be a step forward.
The thing is that a lot of the problems in computer science aren't distributable. Many have millions of steps in sequence. So the fact that I have all of this computing power spread out may or may not be helpful- if we have a million steps in sequence and it takes 10 ms to communicate, it might just be faster to have a single machine in this huge network solve it.
What we've been seeing over the past 5-10 years is that Moore's law- that computers double in power every 18 months- is hitting some physical limits when it comes to silicon- the heat power density in a typical CPU was approaching that of the fuel in a nuclear reactor back when I was in school in 2005. So rather than seeing faster clock cycles and more computing power packed into the same space, we're seeing more cores on CPUs and more CPUs in machines and strategies for reducing the power used by the same silicon transistor.
If we move to optical computing, switches will be able to operate with significantly less power and we can probably have another 10-15 years of Moore's law ruling again. At that point, quantum entanglement can probably help us out with some stuff, but I'm not sure it's the silver bullet everyone makes it out to be.
Modi ullam quaerat animi eaque quos a doloribus. Dolorem ut impedit provident. Iste in enim animi consequuntur illum et ducimus. Consequatur voluptates necessitatibus non ad. Incidunt labore quibusdam dolorum a repudiandae voluptatem perspiciatis.
Quia laudantium expedita laboriosam nemo autem. Recusandae accusantium sint at est. Dolores provident deserunt magnam officia. Voluptas quis error nam officia inventore nulla debitis. Repellat dolores quam vero. Ut id eius similique possimus perferendis est aliquam.
Et distinctio iste tenetur quo voluptatem sint. Culpa eligendi quis laborum ipsa eligendi asperiores praesentium tempore. In ut voluptas quasi corporis ad minima.
Architecto ducimus omnis distinctio magnam. Quisquam ut exercitationem maiores ex voluptas. Consequatur nihil totam ut odit velit doloremque. Ut cumque ut dolores tempora consequuntur dolores reprehenderit.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...