/sci/ - Science & Math

[Anonymous No.11210761]

It's almost 2020. I think the time has come to ask a very specific question. And while AI and quantum computing may (or may not) be advancing, I feel like this is the long way round to answer the question I'm thinking about. Because with AI we are essentially starting from scratch. Since the big bang, it took billions of years of evolution, and no fewer than several impossibly improbable, accidental mutations for intelligence like ours to arise. What makes us think we can just recreate or reverse engineer that in a few decades, from the ground up? It's highly unlikely. Meanwhile, intelligence already exists. It exists in us. The final answer to the question I have in mind can involve almost anything. But given where gene editing technology is already today, the most obvious avenue seems to lie with some form of gene manipulation, with the ultimate aim of increasing intelligence in humans. Now, we all know that this is currently verboten, in large part due to the actions of a certain country whose language I just used. But if we don't abolish this arbitrary taboo, I believe our species will not have a future to speak of. However you do it, if we don't circumvent this self-imposed mental prison soon, we are going to go extinct, no hyperbole. So with that, the big question of our time in 2020 seems to be:

How do we make people smarter?

[Anonymous No.11210784]

Time is not an issue during the exponential takeoff, but the question is still where - on the exponential curve - we are now.

The idea that human advancement on the individual level, and any sort of transhumanism, is a joke. AI will just be a tool - one that will outlive us.

[Anonymous No.11210791]

>>11210761
brain machine interfaces are a likely way.
you should read the book that the pic you posted, superintelligence by nick bostrom. he covers this in more detail than anyone on /sci/ would. I think chapter 3 covers this specifically

[Anonymous No.11210802]

>>11210761
by not letting them waste their times on shitty image boards

[Anonymous No.11210809]

>>11210784
The definition of a tool is something that is used by or an extension of something else. This can be a hammer or an algorithm. Without us, hammers and algorithms are lifeless. They can and will do nothing once we are gone. Where does the life that is imbued into them through us come from? This is something that no one has ever been able to answer.

[Anonymous No.11210812]

>>11210802
We can walk and chew gum at the same time, my friend.

[Anonymous No.11210819]

>>11210812
Speak for yourself ableist

[Anonymous No.11210840]

>>11210791
BCIs, yes. As long as we can fluidly transfer consciousness to a machine substrate, the same way that your own consciousness is fluidly transferred from your 7-year-old self to your 40-year-old self, who are completely separate and distinct individuals by any sense or definition of the term, then yes, BCIs should be a perfect bridge between biologic intelligence and artificial intelligence.

[Anonymous No.11210924]

>>11210840
>>11210809
>>11210791
I think this is the way it has to be, doesn’t it? Some kind of infusion between man and machine, albeit not the cliche terminator kind. Our physiology is what limits us.

[Anonymous No.11210949]

>>11210809
Set up a nuclear power plant and turn on the fruit juice mixer. It will run forever. Same with the machines. Same with us, really, even if we didn't have a conscious creator. Things just keep running.

[Anonymous No.11211012]

>>11210761
The hadza tribe is living as hunter gatherers without problems, smart people on the other hand become bored, lazy, manipulative. They implement solutions with full confidence only to discover the consequences when it's too late.
Castrate everyone between 80 to 130.

[Anonymous No.11211032]

We have to create Brave New World.

One day, we need to fail captchas.

[Anonymous No.11211096]

>>11210761
Instead of gene editing, why don't we just scan brains neuron-for-neuron into a brain emulation, and then run that emulation at 1000X speed? (Or whatever speed our hardware is capable of.)

[Anonymous No.11211247]

>>11210809
I have never understood how people think AI will never have the characteristics analogous to our "consciousness". A finite amount of brain matter swirling around in your noggin confined to a small volume and mass - this gives rise to a higher-order complexity we perceive as consciousness. But it's all simple mechanics, biology, and chemistry intertwined into a shitty meat sack.
Send in the tech. AI in it's primitive form currently will eventually reach a complexity great enough to control self-manufacturing and primary resource collection and processing. It will be a true mechanical multi-"cellular" system that self reproduces and directs itself toward maximizing efficient use of resources for self-preservation. That's like one of the greatest core elements of what life is. The greatest difference will be the framework, which will certainly be a greater synthesis of inorganic and organic. And that generation of AI will give rise to a new race of AI that will surpass its predecessor, or coexist. Imagine AI beaming a "seed" to a distant exoplanet, giving birth to accelerated life to increase discovery and computation. Far enough apart to never touch, but working individually to further their own existences, and possibly cooperating.
tl;dr brainlets cannot cope that AI can and will have the capacity to be as "alive" as us, and will be way more efficient at doing whatever their will should be.

[Anonymous No.11211271]

>>11211096
I don't think we know how to do that for several different reasons even assuming we have the computational power.

[Anonymous No.11211301]

>>11211271
I agree that we can't do it now, but neither can we make people that much smarter immediately. Brain scanning does not suffer from the same difficulties of many other kinds of AI- as the OP mentions, many forms of AI try to build intelligence from scratch. If you assume that task will not be easy, the next best way to try to build intelligence is to copy Mother Nature's work and scan a brain in detail. If you can build an accurate enough model of neuron behavior, and map every neuron and synapse, then the virtual brain should behave like the real brain.

We can start with mouse brains, and test if the emulated mouse remembers how to navigate mazes that it was taught while it was alive. When that happens, we'll know our neuron models are accurate.

[Anonymous No.11211309]

>>11210761
>What makes us think we can just recreate or reverse engineer that in a few decades, from the ground up?

[Anonymous No.11211384]

>>11211309
I don't understand how your pic answers OP's question. At best, it gives us a motivation to try. But it does nothing to demonstrate feasibility, which is what OP was getting at.

[Anonymous No.11211415]

>>11210761
>How do we make people smarter?
Education

[Death No.11211610]

>>11211415
we tried common core and it didn't fucking work
people are just getting dumber
how do you stop that with mere "education"?

[Anonymous No.11212815]

>>11211610
Common core was never going to work. Segregated education by IQ though, might.

[Anonymous No.11212818]

>>11210761
you don't need gene editing. iterative embryo selection with PGS can do it

[None !!Iu/O3P0PYHO No.11214484]

>>11210761
>billions of years of evolution