/sci/ - Science & Math

>>10177248
yes. I don't fucking get it either, the only intelligence we know of is based on chemicals and these retards think the optimal way of recreating it is using a fucking piece of metal.

>>10177267
Only communications we knew was glyphs and sounds and somehow these retards made it into piece of metal, and intelligence doesn't go far beyond what you can communicate, maybe it's you who is retarded.

>>10177267
yeah maybe thats why they fail so much. 70 years of research billions of dollars and the state of the art is a computer that can play a board game...

>>10177345
You're missing out really important stuff, there is computer that can make a voice call to book you a meeting at hairdresser.

>>10177248
Biology doesn't understand intelligence very much.

Matematicians: hold our beer.

>>10177248
Yeah dudeeeee... like the human brain... really crazy stuff uh? Almost like magic bro

>>10177347
>Biology doesn't understand intelligence very much.
Exactly the reason why all the funding should have gone into neuroscience in the first place.

>>10177347
>really important stuff
>book you a meeting at a hairdresser

t. computer scientists

>>10177248
It is because they spite nature, organic functions, and everything that cannot fit exactly into objective, never-changing functions. They wish to replace God with himself, to have an ultimate control over each and every aspect of that which can be considered "life". I am led to believe this is due to the extreme autism and anti-piety required to flourish in higher levels of computatuonal theory.

>>10177343
maybe I actually really am retarded, because I didn't get your post at all.

>>10177267
>metal isn't a chemical

>>10177377
I meant based on chemical transmission as opposed to electric and I admit I phrased it as an idiot

>>10177372
so its because they can't get laid?

>>10177248
I don't think biology even has it covered. Consciousness is poorly understood in its role of intelligence.

>>10177248
>engineers with no education in avian biology want to recreate flight

>>10177267
Chemicals that are only used to induce a transfer of electron potential.

About half of the PhDs at DeepMind are Neuroscientists

>>10177267

computers were invented to replicate abstract human thought

>>10177373
Not him, but communication also didn't start as technology, yet here we are.

Perhaps computers etc are not the optimal way, but science often is less often about that.
We are pushing our limits on all the fields, our chemistry is clearly further away from intteligence than our computer systems are.

>>10177472
No.
Neurotransmitters are not just there to carry (or attenuate) the signal from neuron A to neuron B. It is just as crucial to the function as any other part.
You could just as well say the electric potential is only there to induce the influx of calcium at the synapse.
Neurotransmitters don't just magically tell the cell downstream to fire and then disappear.

If we replaced every chemical synapse in your brain with a direct electrical synapse, you would be dead.

no one has a fucking clue concerning consciousness, if they say so, they are lying out their ass
right now, we understant the core of stars billions of light years away better than the thing that makes us able to know about them
science sure is one hell of a ride

I guess what I mean is, transmitters and receptors don't just carry the signal, they modulate it.
And they don't just modulate it, they modulate it conditionally based on thousands of weird, chemical parameters that we don't even fully understand in nature yet. The system isn't static at all, it's constantly upregulating and downregulating and reuptaking and shitting out gases that synapse in reverse. And all other manner of weird chemical bullshit.

I'm not saying computers can't ever emulate this, but modeling our thought process with simple electrical synapses is never going to get us far.

>>10177248
No, it is because they are arrogant. The probability that a cell gets created randomly out of acids is about as high as a microchip randomly being created out of sand and rare earth laying close to each other. Only God can create life and consciousness, no human can and will ever be able to.

>>10177539
there, i bit
here's your fucking (you), happy?

>>10177267
>he thinks evolution creates systems in an optimal way

>>10177539
Based and redpilled

>>10177248
People in the research field don’t want to recreate human intelligence. The idea is to study the forms of intelligence (defined as the ability to perform tasks, tasks defined as computable processes for which intelligence decides on a finite series of steps until it can halt and the problem is solved) that aren’t necessarily tied to human intelligence. Evolutionary and biology inspired solutions lead to things like neural nets and genetic algorithms, but the study of AI at large (machine learning or otherwise) concerns the ability to accumulate experience and integrate it into a decision making system that is not necessarily tied to the way humans do it.

t. PhD student who hates machine learning

>>10177364
Neuroscience is easier with good machines.

>>10177365
It's important step in future of AI.

>>10177560
>the vertebrate eye isn't optimal
Having our photoreceptors in the back is a feature, not a bug.
In vertebrate eyes, the photoreceptors are right on top of the pigmented epithelium, so they get more light.
Ever wondered why many albinos are legally blind? Melanin absorbs light. The RPE is a major part of what makes our eyes work at all. Octopuses do not and cannot have this layer with their backwards setup.

Our photoreceptors are also closer to the choroid blood supply, which is the main source of nutrients and cleanup for the eye. Birds exclusively get blood from here (and a lot of it) which is partially where we get "eagle eyes" from.
Octopuses have to rely on the slower supply of blood from the back of the retina, through all the nerves, all the way to the photoreceptors. It's a design flaw in the octopus eye that limits nutrient flow to the most important structure.

And there is no visual impairment caused by the "inversion". The choroid layer is transparent, and the cortex is developed to account for it. In fact, some of the glial cells here actually guide light to improve vision here (muller cells actually funnel light into rods and cones to enhance them).
And neither the blind spot nor the blood vessel shadow impact fitness.
Also, we have cones and they don't (not an inversion issue, but it should be mentioned).

There are no downsides to this setup.
Vertebrate eyes are not backwards, they are exactly the way they need to be. The octopus eye trades acuity for simplicity (a feature in and of itself).