/sci/ - Science & Math

[Anonymous No.6642269]

dear people not studying CS, CE, EE or DNA or neuroscience,
why do you continue to stay in your field? Your job will be automated and the singularity is going to happen in our lifetime.
When the singularity happens the computer will be 100000x smarter than any human being. Why do you not devote your life to uploading the human brain into a computer?

[Anonymous No.6642271]

>>6642269
>singularity
>happening in our lifetime

top kek

[Anonymous No.6642272]

>muh singularity

[Anonymous No.6642273]

You know what's also gonna be automated after the singularity, OP? Your low quality shitposting. That's right. Even a poorly written chatbot can produce higher quality bait than you and thus renders your pathetic NEET existence useless. Time to kill yourself.

[Anonymous No.6642278]

The jobs that are the MOST likely to be made redundant are the ones you just named.

[Anonymous No.6642285]

>>6642273
sorry about your job being automated breh

I'm not even trolling with this thread, Im being 100% serious. Any job that can be done by a computer will be done by a computer in the future. Further, if you're so smart and doing research outside of neuroscience/CS, your research will be like 1+1 = 2 when an artificially intelligent being reverse engineers its own brain to become much smarter than any human being

[Anonymous No.6642292]

>>6642285
Why are you still wasting your time manually typing shitposts when a computer could do the same job more effectively? Oh yea, right, it's because you're a NEET. Please kill yourself.

[Anonymous No.6642295]

>>6642269
Moore's law is almost certain to fail entirely in the near future. And I have serious doubts on the hard-takeoff model of the Singularity.

[Anonymous No.6642303]

>>6642269
Because i can program better than the CS majors even though im doing physics and math
>do CS major's course with no programming experience
>come in top 5% of class
im not even good at physics

[Anonymous No.6642306]

>>6642295
Moore's law breaks down when you reach the theoretical limit of storing a byte of information in each atom and processing data at the speed of light. It's not even worth worrying about that because by the time our technology is that sophisticated, the 'singularity' would have already happened.

[Anonymous No.6642310]

>>6642295
We have the computing power to reverse engineer the brain, we just need to know how the brain really works. Why do you have doubts on the intelligence explosion?

Even so, if we do reverse a brain into a computer, the robot will not have to deal with most problems of biology. If the robot is given electricity, then death, disease and hunger is a thing of the past

[Anonymous No.6642323]

>>6642295
Moore's law has nothing to do with our ability to program robust intelligent systems. It could continue on the same trajectory for another century and still not get us any closer to figuring it out. Singularitards are too stupid to realize this, however.

[Anonymous No.6642335]

>>6642310
It's actually doubtful whether we have the power to do that, and most estimates that say we do have the power to run a human-equivalent intelligence need supercomputers. That sort of hardware is difficult to find and expensive to build.

Also, because the brain is so hard to figure out, I find it likely that extending and augmenting it is likely to be a fairly difficult problem as well. It's quite possible that AI intelligence may grow linearly, not exponentially. And if Moore's Law shuts down, expanding intelligence to superhuman levels would require an equal expansion in hardware - hardware which is expensive and takes human timescales to construct.

Also, without fundamental advances in hardware, robots will have problems of their own. For instance, power demands - humans are absurdly, incredibly energy efficient. The human brain takes just 17 watts to run, and our muscles are likewise far more efficient, precise, strong, fast, light, and compact than any known actuator. (Obviously there are actuators that are better in any given one of these categories, but none that come even close in all of them.) And chemical power such as food is vastly superior in energy density than batteries.

The robot's gonna need a LOT more electricity than the human needs food, barring multiple-orders-of-magnitude improvements in energy per FLOPS and actuator efficiency.

[Anonymous No.6642348]

>>6642335
thanks for the response

With IBMs atom moving technology, moores law will probably not shut down. And with the same technology, a 3d printer that can print hardware can probably be made.As far as energy requirements, I feel like a safe nuclear reactor will be made in the next couple decades, and with private space companies going to space, we can launch the waste into deep space

If we can reverse engineer the brain, I am sure the brain that is uploaded could artificially expand in size to exponentially larger to become much more intelligent than any human on the planet.

Time and the free market will tell, I guess

[Anonymous No.6642358]

>>6642348
It's not the ability to make small parts that's really inhibiting Moore's law. Quantum effects and the limits of silicon make scaling down transistor size increasingly difficult, and massive efficiency / waste heat problems make increasing transistor speed or stacking 3-dimensional chips near-impossible.

[Anonymous No.6642361]

Singularity could be any time now, faggots:

http://theinstitute.ieee.org/people/students/student-member-develops-device-that-mimics-brain

[Anonymous No.6642373]

>>6642358
You know we haven't just been scaling stuff down, right?

Moore's Law has proceeded by thousands and thousand of inventions that were necessary to make things work on smaller scales. If it was just a matter of minor refinements to get this far, we'd have jumped straight down to what we've got now back in 1980 or so.

People are already working on quantum dots, single-electron transistors, superconductor logic, ballistic deflection transistors, nanotube logic elements, etc.

You'll see terahertz clock speeds, petabyte RAMs, and billion-core processors in your lifetime. Power consumption can still go way down.

[Anonymous No.6642384]

>>6642373
You realize that none of those are near hitting the market yet, right?

I'm not expecting computers to stop improving. Just for them to stop improving exponentially.

[Anonymous No.6642389]

>>6642384
You realize that they'd do much more than double current performance, right?

[Anonymous No.6642393]

>>6642389
you realize that you're that faggot, right?

[Anonymous No.6642401]

>>6642393
assblast 2 the electric bogaloo

[Anonymous No.6642403]

>>6642389
Sure. I just don't expect them to much more than double performance AND be developed quickly enough AND be cheap enough that cost per GFLOPS drops anywhere near half the cost per 18 months, on average.

[Anonymous No.6642413]

>>6642403
it will, its never been wrong before

[Anonymous No.6642433]

>>6642413
Actually, Moore's Law has been beginning to stutter in the past few years. The next "generation" - Moore's-Law step - has been delayed, and performance gains per generation weren't all that great this time either.

There's only so much further we can go with silicon, and having to bring new technologies to silicon's level of maturity introduces whole new kinds of delays and costs.

[Anonymous No.6642465]

>>6642269

Your forgetting that, if it happens it will change the way people perceive reality. We're taking about huge perspective changes, in terms of the way we perceive ourselves, the world at large and the universe.

The ability to absorb, retain and understand vast quantities of information will inevitably lead dramatic changes in individuals and thus society. As the society changes, so will things like Economics, Governments and etc.

[Anonymous No.6642484]

>>6642384
>Just for them to stop improving exponentially.
Yet they've been improving exponentially for 120 years (back when they weren't called computers yet). pic related

I find it more likely we will find a way to keep the exponential trend going when conventional integrated circuits reach their limits, for instance going to 3D circuits

[Anonymous No.6642518]

>>6642269

>By 2010 computers will disappear. They'll be so small, they'll be embedded in our clothing, in our environment. Images will be written directly to our retina, providing full-immersion virtual reality, augmented real reality. We'll be interacting with virtual personalities.

[Anonymous No.6642544]

>>6642518
This looks shopped, I can tell from some of the pixels and from having seen a few shoops in my life

Also http://www.kurzweilai.net/images/How-My-Predictions-Are-Faring.pdf

The vast majority of his predictions were correct. We have computers in our clothings (iPods), virtual reality is pretty much coming (Oculus Rift), virtual personalities are far from ubiquitous but there are some (Siri)

[Anonymous No.6642797]

>>6642544

>Also http://www.kurzweilai.net/images/How-My-Predictions-Are-Faring.pdf

>The vast majority of his predictions were correct.

Let's ask Ray Kurzweil about how accurate Ray Kurzweil's predictions about the future were.

http://spectrum.ieee.org/computing/software/ray-kurzweils-slippery-futurism