-2
$\begingroup$

As you can see, the USB stick is extending from the possessed robot's finger. The USB stick extends its way straight to the person. If it touches the person, then it'll download the person's consciousness. The USB stick must have been modified in order to do such advanced tasks. If an AI could use a universal serial bus to download information from a person's brain, then the universal serial bus must be modified to be able to connect to neurotransmitters because data will not be able to be extracted unless you are able to detect all electrical signals in the brain. There's a tiny CAT scan within the USB.

Now I believe the process of downloading someone's brain could go in three ways:

  1. The consciousness is copied and the victim did not get harmed(brain scanning).
  2. The consciousness is copied and the victim loses its consciousness (daydream). The brain must have been modified.
  3. The consciousness is copied and not only the victim's consciousness is stored in the digital world, but all of his body particles are stored elsewhere by nanites (going into a game theory).

I want to clarify if these three ways are possible.

enter image description here

$\endgroup$

closed as off-topic by Mołot, 011358 smell, bukwyrm, Ender Look, Frostfyre Feb 19 at 13:27

This question appears to be off-topic. The users who voted to close gave these specific reasons:

If this question can be reworded to fit the rules in the help center, please edit the question.

  • $\begingroup$ Yes, I know this is Science Fiction, that is why I'm not being expected to be 100% realistic. $\endgroup$ – Anonymous Inquirer Jul 5 '17 at 12:35
  • 4
    $\begingroup$ Questions you might be interested in: When will uploaded minds be a reality?, Human to robot interface for data exchange, How much memory is needed to record a human thought $\endgroup$ – Secespitus Jul 5 '17 at 12:44
  • 1
    $\begingroup$ We currently know next to nothing about how the consciousness actually works. So we have no clue, how it could be copied. Everything, no matter if it's a highly complex procedure that requires a huge lab and months of time or touching somebody with an usb stick, is completely random there. $\endgroup$ – Florian Schaetz Jul 5 '17 at 12:45
  • 2
    $\begingroup$ The human brain also contains anywhere from 0.1 to 2 petabytes of information so even if you could somehow download it, it would be incredibly slow. $\endgroup$ – John Jul 5 '17 at 14:54
  • $\begingroup$ Your series of questions about AI is generally poorly received. Most of your questions closed, total score below zero. You may be on your way to question ban. Please, rethink the way you ask. $\endgroup$ – Mołot Jul 5 '17 at 16:46
9
$\begingroup$

No. All three ways are implausible, improbable and, quite likely unscientific. Downloading or uploading minds, consciousness or neurological whatever are based on the fallacy that brains are effectively computers. They aren't, so end of story.

Except. A few additional details. Brains don't have compatible interfaces with USB sticks to permit a transfer of anything. Computers are electronic systems and devices. Brains are squidgy of gelatinous lumps of living stuff with its physiology, biochemistry and electrophysiological signalling. The brain isn't a piece of electronic machinery. Consciousness isn't software. [Aside: even if it was, you bet it wouldn't be PC or Mac compatible. Sorry, just joking. :)]

To transfer someone's consciousness into a computer you would need to build an electronic analogue of that person's brain. It would have to be capable of simulating all the neurological processes (biochemistry, physiology, neural structures, etc etc.) Copying memories into this brain-analogue would be a non-trivial exercise. This apart from all the other non-trivial exercises involved in building a brain-analogue.

Sorry, no, so it's back to the drawing board to devise a better way of uploading minds into machines. Taking into account, this may not be remotely possible.

$\endgroup$
  • 1
    $\begingroup$ I think Linus Torvald's brain might be Linux compatable. And I think Donald Knuth is actually just a TeX document that got out of control while he was writing it! $\endgroup$ – Cort Ammon Jul 5 '17 at 17:33
  • $\begingroup$ The history of computing is riddled with concepts borrowed from how we think our brains work. The neural networks used by the major tech companies are inspired by how our neurons behave. We can't host a human consciousness on a machine yet, but that doesn't mean we won't ever be able to. The existence of AI would prove it's possible to have human level consciousness in a machine. $\endgroup$ – jorfus Jul 5 '17 at 19:10
  • $\begingroup$ @jorfus Human level machine intelligence may eventually be possible. But transferring human minds into machines is more difficult than usually recognized. Possibly, simulating someone's mind might be possible, but that wouldn't be uploading. It's possible to have fun working out the parameters of the technology needed for practical uploading, then realizing how far we from achieving it. $\endgroup$ – a4android Jul 6 '17 at 2:39
  • $\begingroup$ Agreed. What is likely possible is creating a facsimilie of someone's brain at a given point, or recording all the data points that make up someone's mind at a given point in time. It's a bit like Star Trek transporters. Record a being's state, vapourise said being, reconstitute a clone of said being at the other end. $\endgroup$ – Ynneadwraith Feb 19 at 12:36
  • $\begingroup$ @Ynneadwraith Yes. Simulating a facsimile of someone's mind might be possible. Although this isn't uploading. I never liked that explanation of Star Trek transporters. Disintegrate, record, transmit & replicate. There's too much information involved. In principle, it might be possible to make facsimiles so close it would be very hard to distinguish them from their originals. Counterfeit minds! Now's there's a novel idea. Thanks for your thoughtful comment. $\endgroup$ – a4android Feb 20 at 1:26
4
$\begingroup$

Nope.

Complexity

Brains work through a variety of communications methods. Some are electrical, some are chemical. To digitize a brain, you'd basically need to make a CT scan of the target's brain at such a high resolution that it captures all of the brain's neurons in their 3D relation to each other and the exact position of all neurotransmitters in relation to the neurons that are in motion at the time of the scan. And this scan would need to happen very quickly, since the entire 3D map would be changing constantly. If the scan took more than a second or two, your scan has probably introduced too much corruption to be usable.

There's no way to build a USB-enabled device that scan the brain, map out all of those biological components in a 3D space, and then transmit them. Your brain scanner has to map out every molecule in your target's head, convert each molecule to a digital representation of what the molecule is and where it is, and then store it somehow until it can be sent through your USB port.

Size of your data

I do not have the means to give an accurate picture of how much data this brain-scan comprises. But to put it into perspective, this question says human DNA occupies roughly 725 MB of data. Nature says there are roughly 86 billion neurons in the average human brain. This link says DNA is roughly 0.9% of cell weight. Some back-of-the-envelope extrapolation from this tells me that the brain is going to be at least 6,300 Exabytes of data. Probably more.

Bandwidth

Another issue you face is bandwidth. Currently, USB is at version 3.1. A real-world test showed that it takes about 76 seconds to copy 36.7 GB via USB 3.1. At these speeds, it would take you 443,957 years to copy off your brain scan via USB.

What you need

You're going to need to invent a long chain of new technologies here.

  • Scanners that can do molecular-scale identification instantly in real-time with exact internal location tracking of each molecule.
  • Computing power and algorithms to compress your data from exobyte scale down to something manageable, like a few GB at most.
  • Data transfer protocols far superior to existing USB to send the massive data quickly.
  • All of this on a scale that is portable.
  • Oh, and is able to do the scans without destroying the target brain. I mean, unless destroying the target is okay...
$\endgroup$
  • $\begingroup$ I can see that it is a super complexed idea, meaning a lot of things to do in order for it to work in some kind of way. Of course, the USB has to be futuristically modified(not talking about the current ones we have). $\endgroup$ – Anonymous Inquirer Jul 6 '17 at 7:04
  • $\begingroup$ As I might have mentioned, I am not trying to be too realistic, though I regard your ideas on this topic. I understand that no one has actually done it in real life. It is hard to think about something that we don't currently have, but there should be some answers that have a small connection between science fact and science fiction. In other words, I don't use 100% reality when it comes to these futuristic concepts. So when it comes to this sci-fi topic, I tend to find the most reasonable possibilities, rather than the exact answers from reality. $\endgroup$ – Anonymous Inquirer Jul 6 '17 at 7:10
  • $\begingroup$ not looking for 100% magic answers either $\endgroup$ – Anonymous Inquirer Jul 6 '17 at 7:11
  • $\begingroup$ @CaM In the first part of your answer you're talking about some kind of chemical and electrical snapshot of the neurol network, which I agree with. So I feel you lost yourself in the second part, where you talk about DNA which doesn't have a lot to do with anything.(Moreover, if you want to copy the DNA you just need to find one cell, not all the billions) $\endgroup$ – Echox Feb 19 at 8:25
  • $\begingroup$ @Echox, I was trying to find some way to show HOW MUCH data is involved. DNA is a component of cells that I could find hard numbers for. Since it's just one part of the cells, then we can use that to guess at the scale of the total data. Because you need ALL THE MOLECULES by type and 3D location. The data will be exponentially larger than the DNA. $\endgroup$ – CaM Feb 19 at 13:41
3
$\begingroup$

Don't fixate on the USB stick. It's a known technology with known limitations; it won't age well. Nobody has implanted neural link in their head yet so as standards progress over time the brain computer interface will look nothing like USB. Also the human brain won't fit on current portable storage devices. Even with shortcuts and good compression you'd need a shipping container full of drives using current storage tech.(not to mention the computers to run them).

BUT storage density is increasing at a phenomenal rate so it's easy to imagine a future with some very dense storage technologies (maybe 3D optical storage on the myriad internal facets of a crystal --or even more radical single bit single atom techniques)

AND Another limitation is that we currently have no idea how the brain stores data. We don't know the encoding language, we haven't located the storage locations, we have no idea how to read or write to the brain, but we'll figure it out if we don't kill ourselves first. Neural interfaces will be invented, after which some reckless souls will implant them. These would be easy prey. Malicious AI could also invent a method for a rapid, forced neural link. (though the process would almost certainly be painful, invasive and potentially damaging to the host). I can't imagine being able to do it without long tendrils snaking throughout the brain (the brain has no convenient bus or central architecture to plug into). So a forced neural link might look like a metallic squid that snakes around the skull and embeds deep in the brain.

In traditional copying the original data persists. That's actually kind of required to get a perfect copy. But we have no idea how a neural link will function. Maybe it can only copy circuits after they've been activated. (you re-experience your life in moments as the AI forcefully activates every circuit in your head) Maybe a rapid copy literally burns out the circuits as it copies them (look up a video of a lightbulb with the glass broken off of it, it lights up bright and burns out in a puff of smoke), destroying each section of the brain as it copies it. Maybe the AI only needs the personality so the personality, memories, hopes and dreams get burnt out and a drooling husk remains. You can basically have it work any way you like since it hasn't been invented yet and we have no idea how it will work.

$\endgroup$
1
$\begingroup$

Check out this fun Scientific America Article here. Essentially, the human brain is currently thought to hold 2.5 Petrabytes. A USB could store about a Gigabyte of information, which would equal 2.5 million USBs to store one human brain. If one had a DVR of equal storage size, one would need to have it continuously record for 300+ years to totally fill that storage device. Compared to the DNA storage, which is much similar. This guess also runs on the fallicy that human memory = digital memory, which it does not. It's presently unknown how much digital information a memory in the human brain contains and if any compression happens. This only speaks to the data storage capacity an organic brain is capable of, not the space of consumption. Either way, you'd need a pretty big computer to keep all of it intact.

$\endgroup$
0
$\begingroup$

First, it's important that you accept that not current technology will, at any moment, be able to do what you want. So forget real USB adapter, forget CAT scans. It will only look like a USB stick and will do a Zombolium Scan (E.T stuff).

Then you need to acknowledge that the mind of a person isn't just his brain or his electrical patterns. You'll need to get to the whole neural network, which extends in the spine and even around the gut. You'll both need a copy of the electrical and chemical configurations and be able to reproduce the network as well.

A little analogy : If you only know the positions of all cars in the world, you won't be able to foresee where they'll be next. You will also need to know where are the roads and what are the speed limits. It's even better if you can analyze the circulations for a few months to get a good feel of the usal patterns. (End of analogy)

Sooo, for your problem :

  • The Robot starts a simulation of a human neural network in his computer mind.

  • Your fake USB stick connect to a person and proceed to do a whole Zombolium scan. It send waves of Zombol through the body for a few seconds/minutes/hours and from it can get a good snapshot of the whole network, how it is and how it behaves.

  • The snapshot is used to customize the simulation, reproducing the global template and then placing it in the same electrical and chemical configuration than the target.

  • The robot is able to, in real time, send external input in his simulation and transcribe the simulated human reactions in robot reactions.

It's your decision if the Zombolium is harmful or not to the humans. But as it is a lengthy process, the human may need to be deactivated first (but not too hard, he must stay alive).

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.