Live Forever: Uploading the Human Brain

By Raymond Kurzweil, Ph.D.
(The following article appeared at in February 2000. It is here without permission of the author due to publisher's corrupt archive)

Thought to Implant 4: OnNet, please.

Hundreds of shimmering thumbnail images mist into view,  spread fairly evenly across the entire field of pseudovision.

Thought: Zoom upper left, higher, into Winston's image.

Transmit: It's Nellie. Let's connect and chat over croissants.  Rue des Enfants, Paris in the spring, our favorite table, yes?

Four-second pause.

Background thought: Damn it. What's taking him so long?

Receive: I'm here, ma chêre, I'm here! Let's do it!

The thumbnail field mists away, and a café scene swirls into  place. Scent of honeysuckle. Paté. Wine. Light breeze. Nellie  is seated at a quaint table with a plain white tablecloth. An  image of Winston looking 20 and buff mists in across from her.  Message thumbnails occasionally blink against the sky.

Winston: It's so good to see you again, ma chêre! It's been  months! And what a gorgeous choice of bodies! The eyes are  a dead giveaway, though. You always pick those raspberry  eyes. Trés bold, Nellita. So what's the occasion? Part of me is  in the middle of a business meeting in Chicago, so I can't dally.

Nellie: Why do you always put on that muscleman body,  Winston? You know how much I like your real one. Winston  morphs into a man in his early 50s, still overly muscular.

Winston: (laughing) My real body? How droll! No one but my  neurotechnician has seen it for years! Believe me, that's not  what you want. I can do much better! He fans rapidly through a  thousand images, and Nellie grimaces.

Nellie: Damn it! You're just one of Winston's MI's! Where is the  real Winston? I know I used the right connection!

Winston: Nellie, I'm sorry to have to tell you this. There was a  transporter accident a few weeks ago in Evanston, and…  well, I'm lucky they got to me in time for the full upload. I'm all of  Winston that's left. The body's gone.

When Nellie contacts her friend Winston through the Internet  connection in her brain, he is already, biologically speaking,  dead. It is his electronic mind double, a virtual reality twin, that  greets Nellie in their virtual Parisian café. What's surprising  here is not so much the notion that human minds may someday  live on inside computers after their bodies have expired. It's the  fact that this vignette is closer at hand than most people  realize. Within 30 years, the minds in those computers may just  be our own.

The history of technology has shown over and over that as one  mode of technology exhausts its potential, a new more  sophisticated paradigm emerges to keep us moving at an  exponential pace. Between 1910 and 1950, computer  technology doubled in power every three years; between 1950  and 1966, it doubled every two years; and it has recently been  doubling every year.

By the year 2020, your $1,000 personal computer will have the  processing power of the human brain-20 million billion  calculations per second (100 billion neurons times 1,000  connections per neuron times 200 calculations per second per  connection). By 2030, it will take a village of human brains to  match a $1,000 computer. By 2050, $1,000 worth of  computing will equal the processing power of all human brains  on earth.

Of course, achieving the processing power of the human brain  is necessary but not sufficient for creating human level  intelligence in a machine. But by 2030, we'll have the means to  scan the human brain and re-create its design electronically.

Most people don't realize the revolutionary impact of that. The  development of computers that match and vastly exceed the  capabilities of the human brain will be no less important than  the evolution of human intelligence itself some thousands of  generations ago. Current predictions overlook the imminence  of a world in which machines become more like  humans-programmed with replicated brain synapses that  re-create the ability to respond appropriately to human  emotion, and humans become more like machines-our  biological bodies and brains enhanced with billions of  "nanobots," swarms of microscopic robots transporting us in  and out of virtual reality. We have already started down this  road: Human and machine have already begun to meld.

It starts with uploading, or scanning the brain into a computer.  One scenario is invasive: One very thin slice at a time,  scientists input a brain of choice-having been frozen just  slightly before it was going to die-at an extremely high speed.  This way, they can easily see every neuron, every connection  and every neurotransmitter concentration represented in each  synapse-thin layer.

Seven years ago, a condemned killer allowed his brain and  body to be scanned in this way, and you can access all 10  billion bytes of him on the Internet. You can see for yourself  every bone, muscle and section of gray matter in his body. But  the scan is not yet at a high enough resolution to re-create the  interneuronal connections, synapses and neurotransmitter  concentrations that are the key to capturing the individuality  within a human brain.

Our scanning machines today can clearly capture neural  features as long as the scanner is very close to the source.  Within 30 years, however, we will be able to send billions of  nanobots-blood cell-size scanning machines-through every  capillary of the brain to create a complete noninvasive scan of  every neural feature. A shot full of nanobots will someday allow  the most subtle details of our knowledge, skills and  personalities to be copied into a file and stored in a computer.

We can touch and feel this technology today. We just can't  make the nanobots small enough, not yet anyway. But  miniaturization is another one of those accelerating technology  trends. We're currently shrinking the size of technology by a  factor of 5.6 per linear dimension per decade, so it is  conservative to say that this scenario will be feasible in a few  decades. The nanobots will capture the locations,  interconnections and contents of all the nerve cell bodies,  axons, dendrites, presynaptic vesicles, neurotransmitter  concentrations and other relevant neural components. Using  high-speed wireless communication, the nanobots will then  communicate with each other and with other computers that  are compiling the brain-scan database.

If this seems daunting, another scanning project, that of the  human genome, was also considered ambitious when it was  first introduced 12 years ago. At the time, skeptics said the  task would take thousands of years, given current scanning  capabilities. But the project is finishing on time nevertheless  because the speed with which we can sequence DNA has  grown exponentially.

Brain scanning is a prerequisite to Winston and Nellie's virtual  life-and apparent immortality.

In 2029, we will swallow or inject billions of nanobots into our  veins to enter a three dimensional cyberspace-a virtual reality  environment. Already, neural implants are used to counteract  tremors from Parkinson's disease as well as multiple sclerosis.  I have a deaf friend who can now hear what I'm saying because  of his cochlear implant. Under development is a retinal implant  that will perform a similar function for blind people, basically  replacing certain visual processing circuits of the brain.  Recently, scientists from Emory University placed a chip in the  brain of a paralyzed stroke victim who can now begin to  communicate and control his environment directly from his  brain.

But while a surgically introduced neural implant can be placed  in only one or at most a few locations, nanobots can take up  billions or trillions of positions throughout the brain. We already  have electronic devices called neuron transistors that,  noninvasively, allow communication between electronics and  biological neurons. Using this technology, developed at  Germany's Max Planck Institute of Biochemistry, scientists  were recently able to control from their computer the  movements of a living leech.

By taking up positions next to specific neurons, the nanobots  will be able to detect and control their activity. For virtual reality  applications, the nanobots will take up positions next to every  nerve fiber coming from all five of our senses. When we want to  enter a specific virtual environment, the nanobots will suppress  the signals coming from our real senses and replace them with  new, virtual ones. We can then cause our virtual body to move,  speak and otherwise interact in the virtual environment. The  nanobots would prevent our real bodies from moving; instead,  we would have a virtual body in a virtual environment, which  need not be the same as our real body.

Like the experiences Winston and Nellie enjoyed, this  technology will enable us to have virtual interactions with other  people-or simulated people-without requiring any equipment  not already in our heads. And virtual reality will not be as crude  as what you experience in today's arcade games. It will be as  detailed and subtle as real life. So instead of just phoning a  friend, you can meet in a virtual Italian bistro or stroll down a  virtual tropical beach, and it will all seem real. People will be  able to share any type of experience-business, social,  romantic or sexual- regardless of physical proximity.

The trip to virtual reality will be readily reversible since, with  your thoughts alone, you will be able to shut the nanobots off, or  even direct them to leave your body. Nanobots are  programmable, in that they can provide virtual reality one  minute and a variety of brain extensions the next. They can  change their configuration, and even alter their software.

While the combination of human-level intelligence in a machine  and a computer's inherent superiority in the speed, accuracy  and sharing ability of its memory will be formidable-this is not  an alien invasion. It is emerging from within our human-  machine civilization.

But will virtual life and its promise of immortality obviate the  fear of death? Once we upload our knowledge, memories and  insights into a computer, will we have acquired eternal life?  First we must determine what human life is. What is  consciousness anyway? If my thoughts, knowledge,  experience, skills and memories achieve eternal life without  me, what does that mean for me?

Consciousness-a seemingly basic tenet of "living"-is  perplexing and reflects issues that have been debated since  the Platonic dialogues. We assume, for instance, that other  humans are conscious, but when we consider the possibility  that nonhuman animals may be conscious, our understanding  of consciousness is called into question.

The issue of consciousness will become even more  contentious in the 21st century because nonbiological  entities-read: machines-will be able to convince most of us that  they are conscious. They will master all the subtle cues that we  now use to determine that humans are conscious. And they will  get mad if we refute their claims.

Consider this: If we scan me, for example, and record the exact  state, level and position of my every neurotransmitter, synapse,  neural connection and other relevant details, and then  reinstantiate this massive database into a neural computer,  then who is the real me? If you ask the machine, it will  vehemently claim to be the original Ray. Since it will have all of  my memories, it will say, "I grew up in Queens, New York, went  to college at MIT, stayed in the Boston area, sold a few  artificial intelligence companies, walked into a scanner there  and woke up in the machine here. Hey, this technology really  works."

But there are strong arguments that this is really a different  person. For one thing, old biological Ray (that's me) still exists.  I'll still be here in my carbon, cell-based brain. Alas, I (the old  biological Ray) will have to sit back and watch the new Ray  succeed in endeavors that I could only dream of.

But New Ray will have some strong claims as well. He will say  that while he is not absolutely identical to Old Ray, neither is  the current version of Old Ray, since the particles making up  my biological brain and body are constantly changing. It is the  patterns of matter and energy that are semipermanent (that is,  changing only gradually), while the actual material content  changes constantly and very quickly.

Viewed in this way, my identity is rather like the pattern that  water makes when rushing around a rock in a stream. The  pattern remains relatively unchanged for hours, even years,  while the actual material constituting the pattern-the water-is  replaced in milliseconds.

This idea is consistent with the philosophical notion that we  should not associate our fundamental identity with a set of  particles, but rather with the pattern of matter and energy that  we represent. In other words, if we change our definition of  consciousness to value patterns over particles, then New Ray  may have an equal claim to be the continuation of Old Ray.

One could scan my brain and reinstantiate the new Ray while I  was sleeping, and I would not necessarily even know about it. If  you then came to me, and said, "Good news, Ray, we've  successfully reinstantiated your mind file so we won't be  needing your old body and brain anymore," I may quickly  realize the philosophical flaw in the argument that New Ray is a  continuation of my consciousness. I may wish New Ray well,  and realize that he shares my pattern, but I would nonetheless  conclude that he is not me, because I'm still here.

Wherever you wind up on this debate, it is worth noting that  data do not necessarily last forever. The longevity of  information depends on its relevance, utility and accessibility. If  you've ever tried to retrieve information from an obsolete form  of data storage in an old obscure format (e.g., a reel of  magnetic tape from a 1970s minicomputer), you understand  the challenge of keeping software viable. But if we are diligent  in maintaining our mind file, keeping current backups and  porting to the latest formats and mediums, then at least a  crucial aspect of who we are will attain a longevity independent  of our bodies.

What does this super technological intelligence mean for the  future? There will certainly be grave dangers associated with  21st century technologies. Consider unrestrained nanobot  replication. The technology requires billions or trillions of  nanobots in order to be useful, and the most cost-effective way  to reach such levels is through self-replication, essentially the  same approach used in the biological world, by bacteria, for  example. So in the same way that biological self-replication  gone awry (i.e., cancer) results in biological destruction, a  defect in the mechanism curtailing nanobot self-replication  would endanger all physical entities, biological or otherwise.

Other salient questions are: Who is controlling the nanobots?  Who else might the nanobots be talking to?

Organizations, including governments, extremist groups or  even a clever individual, could put trillions of undetectable  nanobots in the water or food supply of an entire population.  These "spy" nanobots could then monitor, influence and even  control our thoughts and actions. In addition, authorized  nanobots could be influenced by software viruses and other  hacking techniques. Just as technology poses dangers today,  there will be a panoply of risks in the decades ahead.

On a personal level, I am an optimist, and I expect that the  creative and constructive applications of this technology will  persevere, as I believe they do today. But there will be a  valuable and increasingly vocal role for a concerned movement  of Luddites-those anti-technologists inspired by  early-19th-century weavers who in protest destroyed machinery  that was threatening their livelihood.

Still, I regard the freeing of the human mind from its severe  physical limitations as a necessary next step in evolution.  Evolution, in my view, is the purpose of life, meaning that the  purpose of life-and of our lives-is to evolve.

What does it mean to evolve? Evolution moves toward greater  complexity, elegance, intelligence, beauty, creativity and love.  And God has been called all these things, only without any  limitation, infinite. While evolution never reaches an infinite  level, it advances exponentially, certainly moving in that  direction. Technological evolution, therefore, moves us  inexorably closer to becoming like God. And the freeing of our  thinking from the severe limitations of our biological form may  be regarded as an essential spiritual quest.

By the close of the next century, nonbiological intelligence will  be ubiquitous. There will be few humans without some form of  artificial intelligence, which is growing at a double exponential  rate, whereas biological intelligence is basically at a standstill.  Nonbiological thinking will be trillions of trillions of times more  powerful than that of its biological progenitors, although it will  be still of human origin.

Ultimately, however, the earth's technology-creating species  will merge with its own computational technology. After all, what  is the difference between a human brain enhanced a  trillion-fold by nanobot-based implants, and a computer whose  design is based on high-resolution scans of the human brain,  and then extended a trillion-fold?

This may be the ominous, existential question that our own  children, certainly our grandchildren, will face. But at this point,  there's no turning back. And there's no slowing down.

 Back to Main Features Page