5 ms·

Something I've been thinking about a lot lately: if / when it becomes possible to "upload" our minds, wouldn't that be the ideal way to explore the universe? Sp

by icey 13y ago

Something I've been thinking about a lot lately: if / when it becomes possible to "upload" our minds, wouldn't that be the ideal way to explore the universe? Space travel becomes much less daunting when you don't have to worry about an ordinary human lifespan, along with all the intricate & inefficient fueling and environmental needs a biological organism requires. I hope if there's ever an Kurzweilian singularity that the AIs come with a lust for discovering the universe and spread across it like dandelion seeds.

claudius 13y ago
When. I am willing to bet that neither of us nor any of our grandchildren will be alive when it becomes possible to copy one’s personality onto a robot of sorts.
- geuis 13y ago
  The future isn't something in the future anymore. We are living in the middle of it. I'm 33 and can easily see the effects of rapid technological change since the time I was a boy. Things are just getting faster. Betting against change has been a losing bet for decades, if not arguably centuries. As a practical example: solar power. It's been commercially a available for at least 30 years. It's often been derided and dismissed because even now, it's less than 1% of energy production in the US. 30 years of no significant traction, so an obvious loser right? Here's an Economist article from Dec. http://www.economist.com/blogs/graphicdetail/2012/12/daily-chart-19 http://www.economist.com/blogs/graphicdetail/2012/12/daily-c... The price per watt has been falling logarithmically. This corresponds to a rise in efficiency. If you look at the last 5 years or so, you can see that solar panels are starting to hit the up slope of a hockey stick. Within a very few number of years, be prepared to see an incredible impact on the energy industry. It's not going to make oil and gas obsolete, but we'll see cheap energy being used in lots and lots of places. My personal opinion, not really supported by hard data, is that we'll see an African renaissance in the next 10-20 years. Contrary to Western belief, Africa is not entirely desert and full of starving black kids who's parents are never around for some reason. It's also not entirely in some state of civil war. There's problems, but only in spots. What many African countries are doing is leap frogging into modern times. They are bypassing the massive infrastructures that Western nations have and using modern tech to fix their problems. With the spread of inexpensive solar power, you'll see the darkest night time parts of the continent lighting up.
  - claudius 13y ago
    Uh, I’m not saying the future isn’t now. I am saying that uploading one’s mind (in whatever form that will take place) will require much much more energy and complexity than currently available. In fact, I’d go so far to say that we find a way around relativity before we find a way to remodel the human brain – it is just so mind-boggingly complex, based on such difficult-to-simulate physics, that I don’t think it will be done any time soon. And I fully agree with you on solar power, it is a wonderful thing, though Germany is possibly not the brightest place to put the largest installation in terms of energy output…
    - geuis 13y ago
      You're making the same mistake again. You're equating complexity to impossibility/improbability. They are not causally related. We know that brains work. You and I having are having this disconnected conversation using an incredibly deep stack of technological infrastructure, all created with human brains over decades and centuries. Brains are physical things, not metaphysical. The physics of the universe we live in allow matter to interact in particular ways that let brains exist. The physics that lets brains exist is the same physics that describe relativity. We have 100+ years of thousands upon thousands of experiments and acquisition of data that backs up relativity. Its very unlikely to go anywhere. I'll agree that the physical structure of an adult brain (not just human) is very complex. Its not mind boggling though. Much of the brain is just repeated neural structures. There are what, a few trillion connections? In 1950 that was a mind boggling huge amount of data. Now? That amount of data fits into a thumb drive with room to spare. The tricky part up to now, and a couple years into the future, is building the right tools to pick it all apart. Its just like the solar panel problem. Low efficiencies, high cost. Same with Moore's Law and cpu's. But we have, if not the highly refined eventual versions, but the basic and functional tools to brute force through figuring out the its & bits of the brain. Actually, a lot of the work is already done. Neuro chemistry, effects of different neurotransmitters, effects of various drugs on the brain, FMRI, etc and etc. Don't forget that Obama recently announced a brain initiative. Its the along the same line as the human genome project in the 1990's. Detractors said that it would take hundreds of years to finish with technology in the early 90's. Was done under budget and on time within 10 years. By 2020, just 7 short years away, everything computer-related is going to have doubled a little over 3 more times. With all of this in place, the subsystems of the brain are complicated but not insurmountably. Even without initially having a sound mathematical description of how the brain works, we are/will have enough data to brute force a detailed simulation. If that's the route we have to go, then the mathematics will come later. Indeed, this is how much of the relationship between engineering and science/math has been for thousands of years. People were using arches in buildings long, long before understanding the math behind weight distribution. In terms of energy, we can easily calculate how much energy it takes to simulate a brain. Let me put some numbers together. I was born 1/30/1980. I am 12,149 days old. Lets say that on average, it has taken 2200 calories a day for me to survive from birth to the current moment. Roughly, it has taken 26,727,800 calories for my body and brain to survive and prosper. To keep this simple, I won't even try to calculate the energy spent on everything else like clean water, electricity, transportation, entertainment, etc. In general terms, since I live on Earth with you and billions of others, its not an insurmountable amount of energy since we're all here. Plus, I'm a big guy so its safe to say it takes less calories to raise a human-class sentience to adulthood. So, lets sum up your arguments and my counters: 1) Uploading (understanding) a brain is too complex: -> There's already a large understanding of the biochemistry and mathematics behind brain function. Its not a mystical black box. Given current and in-development tools, combined with massive data storage and analysis systems, picking apart all the bits of the brain will be done in a few years. 2) Too much energy: -> We're already well on the way to reverse engineering the brain. The energy required to finish the task is marginal. To simulate a brain, we already grow billions of real brains for less calories than it takes for a fat guy in San Francisco to write everything you've been reading. Convert those thermal calories to heat in the world's most efficient sterling engine, and that's one thing you can use to power your simulated brain. I dare say that it will cost less power to run a sentient AI than it does to support my love of bacon. 3) We'll find a way around relativity before understanding the brain: -> Brains exist. So does relativity. The physics of the universe support reverse engineering brains. They don't experimentally support breaking relativity at the moment. However, I too am hopeful that some loophole or something is discovered that lets relativity be less of a barrier than it appears. However, given practical real-world data, that isn't likely any time soon. 4) Physics is difficult to simulate: -> I didn't hit on this point explicitly above, but yes and no. Yes in that creating a simulation of universe from particles on up is daunting. Even with today's and near-future doubling growth, doing massive particle simulations will be very hard. No, in that you don't have to simulate every atom in a brain to understand how it works. You create mathematical models detailing how the tiny little structural bits of the brain work (neurotransmitters for example) and perform calculations based at that. Still a large computational problem, but one that is already with our grasp as of this minute and only easier to do going into the future. So basically, you don't actually have to simulate physics. You just have to simulate the mathematical interactions of the subsystems that make up the brain to get comparable results. 5) Germans -> That's the great thing about solar power. Its everywhere. During the day in most parts of the world, even when its overcast, you still get enough sunlight hitting the ground to provide lots and lots of power. Sure, you'll get more in the desert on the equator, but that's the maximum and not minimum useful amount. People that make the statement you did have the argument backwards. You have to ask, "What is the minimal amount of solar energy we can collect to be useful and economical?" Man, this took a lot longer to write than I thought it would.
columbo 13y ago
> "upload" our minds ... to explore the universe I don't think it'll ever happen. Uploading something just means making a copy, and if that becomes possible the AI potential would be millions of generations smarter than a carbon copy of a human brain. Sort of like digitizing a record or wax cylinder - we don't do it for the quality, we do it for historical or nostalgic purposes. If AI becomes possible then the only reason to upload a human algorithm would be for the same reasons as the wax cylinder. It wouldn't be to rely on the created construct to do decision making or exploring. More likely AI programs would do all the exploring and it would be brought back holodeck/matrix style for people to experience safely in the confines of earth.
- icey 13y ago
  Never? I agree our minds are incredibly complex, but it's not magic. Once humankind can measure what happens inside a brain, it's surely should only be a matter of time before it can replicate its machinery. I agree it may be something that's quite far into the future, but consider how far medicine has come in the past 25, or even 10 years with regards to replacing other human organs. My personal opinion is that AGI will look a lot more like a convergence of man with machine than software suddenly gaining consciousness. I think it will be a viewed as performance enhancement much more than the merging of two foreign entities. I think it will begin with an interface that will allow us to communicate with software neurally, or store our memories externally (even if it's just copies). Early research is already happening in rats to this effect[1][2], I have a hard time believing that our brains are so complex that they will never be deciphered. [1] http://www.technologyreview.com/news/424452/a-first-step-toward-a-prosthesis-for-memory/ http://www.technologyreview.com/news/424452/a-first-step-tow... [2] http://www.technologyreview.com/news/511721/rats-communicate-through-brain-chips/ http://www.technologyreview.com/news/511721/rats-communicate...
  - exit 13y ago
    'columbo' didn't say uploading would never happen, just that uploading in order to explore the universe doesn't make sense
- exit 13y ago
exit 13y ago
you should read the work of "greg egan" in particular his novel "schild's ladder", in which space travel mostly involves digital brains being transmitted by laser between physical outposts sometimes "people" send along specifications for biological hosts in which to be embodied, sometimes not
- quotemstr 13y ago
  You barely beat me to mentioning Egan. I find it interesting that he's able to write engaging stories using AIs as the principal characters, as if to say that most of the attributes we normally attribute to being "human" are actually general properties of sentient minds in general, and we just conflate the two because we happen to be the only sentient minds we know about.
- icey 13y ago
  Thanks for the recommendation, I'm going to be doing some traveling later this week & have been looking for some reading to take along!
  - sbirchall 13y ago
    I cannot recommend Hannu Rajaniemi enough. Mind. Blowing. This goes for the whole HN crowd: read his books.
- cpeterso 13y ago
  Traveling by "AI over laser" is a clever cheat for the speed of light that I hadn't thought of. But how do the physical outputs on the receiving end get there in the first place? <:)
  - DanBC 13y ago
    > Traveling by "AI over laser" I hope the error correction is good. Imagine getting bits of the transmission knocked out by dust.
  - exit 13y ago
    > It has been theorized that a self-replicating starship utilizing relatively conventional theoretical methods of interstellar travel (i.e., no exotic faster-than-light propulsion such as "warp drive", and speeds limited to an "average cruising speed" of 0.1c.) could spread throughout a galaxy the size of the Milky Way in as little as half a million years. http://en.wikipedia.org/wiki/Self-replicating_spacecraft http://en.wikipedia.org/wiki/Self-replicating_spacecraft citing http://www.rfreitas.com/Astro/ComparisonReproNov1980.htm http://www.rfreitas.com/Astro/ComparisonReproNov1980.htm
mtgx 13y ago
Long range communication that travels faster than light should happen sooner than we can make ships that travel faster than light. So if anything, we'll just get video-feeds from there, and just send robots.

by icey 13y ago

claudius 13y ago

When. I am willing to bet that neither of us nor any of our grandchildren will be alive when it becomes possible to copy one’s personality onto a robot of sorts.

geuis 13y ago
The future isn't something in the future anymore. We are living in the middle of it. I'm 33 and can easily see the effects of rapid technological change since the time I was a boy. Things are just getting faster. Betting against change has been a losing bet for decades, if not arguably centuries. As a practical example: solar power. It's been commercially a available for at least 30 years. It's often been derided and dismissed because even now, it's less than 1% of energy production in the US. 30 years of no significant traction, so an obvious loser right? Here's an Economist article from Dec. http://www.economist.com/blogs/graphicdetail/2012/12/daily-chart-19 http://www.economist.com/blogs/graphicdetail/2012/12/daily-c... The price per watt has been falling logarithmically. This corresponds to a rise in efficiency. If you look at the last 5 years or so, you can see that solar panels are starting to hit the up slope of a hockey stick. Within a very few number of years, be prepared to see an incredible impact on the energy industry. It's not going to make oil and gas obsolete, but we'll see cheap energy being used in lots and lots of places. My personal opinion, not really supported by hard data, is that we'll see an African renaissance in the next 10-20 years. Contrary to Western belief, Africa is not entirely desert and full of starving black kids who's parents are never around for some reason. It's also not entirely in some state of civil war. There's problems, but only in spots. What many African countries are doing is leap frogging into modern times. They are bypassing the massive infrastructures that Western nations have and using modern tech to fix their problems. With the spread of inexpensive solar power, you'll see the darkest night time parts of the continent lighting up.
- claudius 13y ago
  Uh, I’m not saying the future isn’t now. I am saying that uploading one’s mind (in whatever form that will take place) will require much much more energy and complexity than currently available. In fact, I’d go so far to say that we find a way around relativity before we find a way to remodel the human brain – it is just so mind-boggingly complex, based on such difficult-to-simulate physics, that I don’t think it will be done any time soon. And I fully agree with you on solar power, it is a wonderful thing, though Germany is possibly not the brightest place to put the largest installation in terms of energy output…
  - geuis 13y ago
    You're making the same mistake again. You're equating complexity to impossibility/improbability. They are not causally related. We know that brains work. You and I having are having this disconnected conversation using an incredibly deep stack of technological infrastructure, all created with human brains over decades and centuries. Brains are physical things, not metaphysical. The physics of the universe we live in allow matter to interact in particular ways that let brains exist. The physics that lets brains exist is the same physics that describe relativity. We have 100+ years of thousands upon thousands of experiments and acquisition of data that backs up relativity. Its very unlikely to go anywhere. I'll agree that the physical structure of an adult brain (not just human) is very complex. Its not mind boggling though. Much of the brain is just repeated neural structures. There are what, a few trillion connections? In 1950 that was a mind boggling huge amount of data. Now? That amount of data fits into a thumb drive with room to spare. The tricky part up to now, and a couple years into the future, is building the right tools to pick it all apart. Its just like the solar panel problem. Low efficiencies, high cost. Same with Moore's Law and cpu's. But we have, if not the highly refined eventual versions, but the basic and functional tools to brute force through figuring out the its & bits of the brain. Actually, a lot of the work is already done. Neuro chemistry, effects of different neurotransmitters, effects of various drugs on the brain, FMRI, etc and etc. Don't forget that Obama recently announced a brain initiative. Its the along the same line as the human genome project in the 1990's. Detractors said that it would take hundreds of years to finish with technology in the early 90's. Was done under budget and on time within 10 years. By 2020, just 7 short years away, everything computer-related is going to have doubled a little over 3 more times. With all of this in place, the subsystems of the brain are complicated but not insurmountably. Even without initially having a sound mathematical description of how the brain works, we are/will have enough data to brute force a detailed simulation. If that's the route we have to go, then the mathematics will come later. Indeed, this is how much of the relationship between engineering and science/math has been for thousands of years. People were using arches in buildings long, long before understanding the math behind weight distribution. In terms of energy, we can easily calculate how much energy it takes to simulate a brain. Let me put some numbers together. I was born 1/30/1980. I am 12,149 days old. Lets say that on average, it has taken 2200 calories a day for me to survive from birth to the current moment. Roughly, it has taken 26,727,800 calories for my body and brain to survive and prosper. To keep this simple, I won't even try to calculate the energy spent on everything else like clean water, electricity, transportation, entertainment, etc. In general terms, since I live on Earth with you and billions of others, its not an insurmountable amount of energy since we're all here. Plus, I'm a big guy so its safe to say it takes less calories to raise a human-class sentience to adulthood. So, lets sum up your arguments and my counters: 1) Uploading (understanding) a brain is too complex: -> There's already a large understanding of the biochemistry and mathematics behind brain function. Its not a mystical black box. Given current and in-development tools, combined with massive data storage and analysis systems, picking apart all the bits of the brain will be done in a few years. 2) Too much energy: -> We're already well on the way to reverse engineering the brain. The energy required to finish the task is marginal. To simulate a brain, we already grow billions of real brains for less calories than it takes for a fat guy in San Francisco to write everything you've been reading. Convert those thermal calories to heat in the world's most efficient sterling engine, and that's one thing you can use to power your simulated brain. I dare say that it will cost less power to run a sentient AI than it does to support my love of bacon. 3) We'll find a way around relativity before understanding the brain: -> Brains exist. So does relativity. The physics of the universe support reverse engineering brains. They don't experimentally support breaking relativity at the moment. However, I too am hopeful that some loophole or something is discovered that lets relativity be less of a barrier than it appears. However, given practical real-world data, that isn't likely any time soon. 4) Physics is difficult to simulate: -> I didn't hit on this point explicitly above, but yes and no. Yes in that creating a simulation of universe from particles on up is daunting. Even with today's and near-future doubling growth, doing massive particle simulations will be very hard. No, in that you don't have to simulate every atom in a brain to understand how it works. You create mathematical models detailing how the tiny little structural bits of the brain work (neurotransmitters for example) and perform calculations based at that. Still a large computational problem, but one that is already with our grasp as of this minute and only easier to do going into the future. So basically, you don't actually have to simulate physics. You just have to simulate the mathematical interactions of the subsystems that make up the brain to get comparable results. 5) Germans -> That's the great thing about solar power. Its everywhere. During the day in most parts of the world, even when its overcast, you still get enough sunlight hitting the ground to provide lots and lots of power. Sure, you'll get more in the desert on the equator, but that's the maximum and not minimum useful amount. People that make the statement you did have the argument backwards. You have to ask, "What is the minimal amount of solar energy we can collect to be useful and economical?" Man, this took a lot longer to write than I thought it would.

columbo 13y ago

> "upload" our minds ... to explore the universe I don't think it'll ever happen. Uploading something just means making a copy, and if that becomes possible the AI potential would be millions of generations smarter than a carbon copy of a human brain. Sort of like digitizing a record or wax cylinder - we don't do it for the quality, we do it for historical or nostalgic purposes. If AI becomes possible then the only reason to upload a human algorithm would be for the same reasons as the wax cylinder. It wouldn't be to rely on the created construct to do decision making or exploring. More likely AI programs would do all the exploring and it would be brought back holodeck/matrix style for people to experience safely in the confines of earth.

icey 13y ago
Never? I agree our minds are incredibly complex, but it's not magic. Once humankind can measure what happens inside a brain, it's surely should only be a matter of time before it can replicate its machinery. I agree it may be something that's quite far into the future, but consider how far medicine has come in the past 25, or even 10 years with regards to replacing other human organs. My personal opinion is that AGI will look a lot more like a convergence of man with machine than software suddenly gaining consciousness. I think it will be a viewed as performance enhancement much more than the merging of two foreign entities. I think it will begin with an interface that will allow us to communicate with software neurally, or store our memories externally (even if it's just copies). Early research is already happening in rats to this effect[1][2], I have a hard time believing that our brains are so complex that they will never be deciphered. [1] http://www.technologyreview.com/news/424452/a-first-step-toward-a-prosthesis-for-memory/ http://www.technologyreview.com/news/424452/a-first-step-tow... [2] http://www.technologyreview.com/news/511721/rats-communicate-through-brain-chips/ http://www.technologyreview.com/news/511721/rats-communicate...
- exit 13y ago
  'columbo' didn't say uploading would never happen, just that uploading in order to explore the universe doesn't make sense
exit 13y ago

exit 13y ago

you should read the work of "greg egan" in particular his novel "schild's ladder", in which space travel mostly involves digital brains being transmitted by laser between physical outposts sometimes "people" send along specifications for biological hosts in which to be embodied, sometimes not

quotemstr 13y ago
You barely beat me to mentioning Egan. I find it interesting that he's able to write engaging stories using AIs as the principal characters, as if to say that most of the attributes we normally attribute to being "human" are actually general properties of sentient minds in general, and we just conflate the two because we happen to be the only sentient minds we know about.
icey 13y ago
Thanks for the recommendation, I'm going to be doing some traveling later this week & have been looking for some reading to take along!
- sbirchall 13y ago
  I cannot recommend Hannu Rajaniemi enough. Mind. Blowing. This goes for the whole HN crowd: read his books.
cpeterso 13y ago
Traveling by "AI over laser" is a clever cheat for the speed of light that I hadn't thought of. But how do the physical outputs on the receiving end get there in the first place? <:)
- DanBC 13y ago
  > Traveling by "AI over laser" I hope the error correction is good. Imagine getting bits of the transmission knocked out by dust.
- exit 13y ago
  > It has been theorized that a self-replicating starship utilizing relatively conventional theoretical methods of interstellar travel (i.e., no exotic faster-than-light propulsion such as "warp drive", and speeds limited to an "average cruising speed" of 0.1c.) could spread throughout a galaxy the size of the Milky Way in as little as half a million years. http://en.wikipedia.org/wiki/Self-replicating_spacecraft http://en.wikipedia.org/wiki/Self-replicating_spacecraft citing http://www.rfreitas.com/Astro/ComparisonReproNov1980.htm http://www.rfreitas.com/Astro/ComparisonReproNov1980.htm

mtgx 13y ago

Long range communication that travels faster than light should happen sooner than we can make ships that travel faster than light. So if anything, we'll just get video-feeds from there, and just send robots.