Sunday, 31 May 2009

Create an AI on Your Computer

Written on May 28, 2009 – 11:48 am | by Aaron Saenz |
Intelligence Realm is seeking to build the first AI using distributed computing.

Intelligence Realm is seeking to build the first AI using distributed computing.

If many hands make light work, then maybe many computers can make an artificial brain. That’s the basic reasoning behind Intelligence Realm’sArtificial Intelligence project. By reverse engineering the brain through a simulation spread out over many different personal computers, Intelligence Realm hopes to create an AI from the ground-up, one neuron at a time. The first waves of simulation are already proving successful, with over 14,000 computers used and 740 billion neurons modeled. Singularity Hub managed to snag the project’s leader, Ovidiu Anghelidi, for an interview: see the full text at the end of this article.

The ultimate goal of Intelligence Realm is to create an AI or multiple AIs, and use these intelligences in scientific endeavors. By focusing on the human brain as a prototype, they can create an intelligence that solves problems and “thinks” like a human. This is akin to the work done at FACETS that Singularity Hub highlighted some weeks ago. The largest difference between Intelligence Realm and FACETS is that Intelligence Realm is relying on a purely simulated/software approach.

Which sort of makes Intelligence Realm similar to the Blue Brain Project that Singularity Hub also discussed. Both are computer simulations of neurons in the brain, but Blue Brain’s ultimate goal is to better understand neurological functions, while Intelligence Realm is seeking to eventually create an AI. In either case, to successfully simulate the brain in software alone, you need a lot of computing power. Blue Brain runs off a high-tech supercomputer, a resource that’s pretty much exclusive to that project. Even with that impressive commodity, Blue Brain is hitting the limit of what it can simulate. There’s too much to model for just one computer alone, no matter how powerful. Intelligence Realm is using a distributed computing solution. Where one computer cluster alone may fail, many working together may succeed. Which is why Intelligence Realm is looking for help.

The AI system project is actively recruiting, with more than 6700 volunteers answering the call. Each volunteer runs a small portion of the larger simulation on their computer(s) and then ships the results back to the main server. BOINC, the Berkeley built distributed computing software that makes it all possible, manages the flow of data back and forth. It’s the same software used for SETI’s distributed computing processing. Joining the project is pretty simple: you just download BOINC, some other data files, and you’re good to go. You can run the simulation as an application, or as part of your screen saver.

Baby Steps

So, 6700 volunteers, 14,000 or so platforms, 740 billion neurons, but what is the simulated brain actually thinking? Not a lot at the moment. The same is true with the Blue Brain Project, or FACETS. Simulating a complex organ like the brain is a slow process, and the first steps are focused on understanding how the thing actually works. Inputs (Intelligence Realm is using text strings) are converted into neuronal signals, those signals are allowed to interact in the simulation and the end state is converted back to an output. It’s a time and labor (computation) intensive process. Right now, Intelligence Realm is just building towards simple arithmetic.

Which is definitely a baby step, but there are more steps ahead. Intelligence Realm plans on learning how to map numbers to neurons, understanding the kind of patterns of neurons in your brain that represent numbers, and figuring out basic mathematical operators (addition, subtraction, etc). From these humble beginnings, more complex reasoning will emerge. At least, that’s the plan.

Intelligence Realm isn’t just building some sort of biophysical calculator. Their brain is being designed so that it can change and grow, just like a human brain. They’ve focused on simulating all parts of the brain (including the lower reasoning sections) and increasing the plasticity of their model. Right now it’s stumbling towards knowing 1+1 = 2. Even with linear growth they hope that this same stumbling intelligence will evolve into a mental giant. It’s a monumental task, though, and there’s no guarantee it will work. Building artificial intelligence is probably one of the most difficult tasks to undertake, and this early in the game, it’s hard to see if the baby steps will develop into adult strides. The simulation process may not even be the right approach. It’s a valuable experiment for what it can teach us about the brain, but it may never create an AI. A larger question may be, do we want it to?

Knock, Knock…It’s Inevitability

With the newest Terminator movie out, it’s only natural to start worrying about the dangers of artificial intelligence again. Why build these things if they’re just going to hunt down Christian Bale? For many, the threats of artificial intelligence make it seem like an effort of self-destructive curiosity. After all, from Shelley’s Frankenstein Monster to Adam and Eve, Western civilization seems to believe that creations always end up turning on their creators.

AI, however, promises rewards as well as threats. Problems in chemistry, biology, physics, economics, engineering, and astronomy, even questions of philosophy could all be helped by the application of an advanced AI. What’s more, as we seek to upgrade ourselves through cybernetics and genetic engineering, we will become more artificial. In the end, the line between artificial and natural intelligence may be blurred to a point that AIs will seem like our equals, not our eventual oppressors. However, that’s not a path that everyone will necessarily want to walk down.

Will AI and Humans learn to co-exist?

Will AI and Humans learn to co-exist?

The nature of distributed computing and BOINC allow you to effectively vote on whether or not this project will succeed. Intelligence Realm will eventually need hundred of thousands if not millions of computing platforms to run their simulations. If you believe that AI deserves a chance to exist, give them a hand and recruit others. If you think we’re building our own destroyers, then don’t run the program. In the end, the success or failure of this project may very well depend on how many volunteers are willing to serve as mid-wives to a new form of intelligence.

Before you make your decision though, make sure to read the following interview. As project leader, Ovidiu Anghelidi is one of the driving minds behind reverse engineering the brain and developing the eventual AI that Intelligence Realm hopes to build. He’s didn’t mean for this to be a recruiting speech, but he makes some good points:

SH: Hello. Could you please start by giving yourself and your project a brief introduction?

OA: Hi. My name is Ovidiu Anghelidi and I am working on a distributed computing project involving thousands of computers in the field of artificial intelligence. Our goal is to develop a system that can perform automated research.

What drew you to this project?

During my adolescence I tried understanding the nature of question. I used extensively questions as a learning tool. That drove me to search for better understanding methods. After looking at all kinds of methods, I kinda felt that understanding creativity is a worthier pursuit. Applying various methods of learning and understanding is a fine job, but finding outstanding solutions requires much more than that. For a short while I tried understanding how creativity is done and what exactly is it. I found out that there is not much work done on this subject, mainly because it is an overlapping concept. The search for creativity led me to the field of AI. Because one of the past presidents of the American Association of Artificial Intelligence dedicated an entire issue to this subject I started pursuing that direction. I looked into the field of artificial intelligence for a couple of years and at some point I was reading more and more papers that touched the subject of cognition and brain so I looked briefly into neuroscience. After I read an introductory book about neuroscience, I realized that understanding brain mechanisms is what I should have done all along, for the past 20 years. To this day I am pursuing this direction.

What’s your time table for success? How long till we have a distributed AI running around using your system?

I have been working on this project for about 3 years now, and I estimate that we will need another 7-8 years to finalize the project. Nonetheless we do not need that much time to be able to use some its features. I expect to have some basic features that work within a couple of months. Take for example the multiple simulations feature. If we want to pursue various directions in different fields (i.e. mathematics, biology, physics) we will need to set up a simulation for each field. But we do not need to get to the end of the project, to be able to run single simulations.

Do you think that Artificial Intelligence is a necessary step in the evolution of intelligence? If not, why pursue it? If so, does it have to happen at a given time?

I wouldn’t say necessary, because we don’t know what we are evolving towards. As long as we do not have the full picture from beginning to end, or cases from other species to compare our history to, we shouldn’t just assume that it is necessary.

We should pursue it with all our strength and understanding because soon enough it can give us a lot of answers about ourselves and this Universe. By soon I mean two or three decades. A very short time span, indeed. Artificial Intelligence will amplify a couple of orders of magnitude our research efforts across all disciplines.

In our case it is a natural extension. Any species that reaches a certain level of intelligence, at some point in time, they would start replicating and extending their natural capacities in order to control their environment. The human race did that for the last couple thousands of years, we tried to replicate and extend our capacity to run, see, smell and touch. Now it reached thinking. We invented vehicles, television sets, other devices and we are now close to have artificial intelligence.

What do you think are important short term and long term consequences of this project?

We hope that in short term we will create some awareness in regards to the benefits of artificial intelligence technology. Longer term it is hard to foresee.

How do you see Intelligence Realm interacting with more traditional research institutions? (Universities, peer reviewed Journals, etc)

Well…, we will not be able to provide full details about the entire project because we are pursuing a business model, so that we can support the project in the future, so there is little chance of a collaboration with a University or other research institution. Down the road, as we we will be in an advanced stage with the development, we will probably forge some collaborations. For the time being this doesn’t appear feasible. I am open to collaborations but I can’t see how that would happen.

I submitted some papers to a couple of journals in the past, but I usually receive suggestions that I should look at other journals, from other fields. Most of the work in artificial intelligence doesn’t have neuroscience elements and the work in neuroscience contains little or no artificial intelligence elements. Anyway, I need no recognition.

Why should someone join your project? Why is this work important?

If someone is interested in artificial intelligence it might help them having a different view on the subject and seeing what components are being developed over time. I can not tell how important is this for someone else. On a personal level, I can say that because my work is important to me and by having an AI system I will be able to get answers to many questions, I am working on that. Artificial Intelligence will provide exceptional benefits to the entire society.

What should someone do who is interested in joining the simulation? What can someone do if they can’t participate directly? (Is there a “write-your-congressman” sort of task they could help you with?)

If someone is interested in joining the project they need to download the Boinc client from the site and then attach to the project using the master Url for this project, We appreciate the support received from thousands of volunteers from all over the world.

If someone can’t participate directly I suggest to him/her to keep an open mind about what AI is and how it can benefit them. He or she should also try to understand its pitfalls.

There is no write-your-congressman type of task. Mass education is key for AI success. This project doesn’t need to be in the spotlight.

What is the latest news?

We reached 14,000 computers and we simulated over 740 billion neurons. We are working on implementing a basic hippocampal model for learning and memory.

Anything else you want to tell us?

If someone considers the development of artificial intelligence impossible or too far into the future to care about, I can only tell him or her, “Embrace the inevitable”. The advances in the field of neuroscience are increasing rapidly. Scientists are thorough.

Understanding its benefits and pitfalls is all that is needed.

Thank you for your time and we look forward to covering Intelligence Realm as it develops further.

Thank you for having me.

Saturday, 30 May 2009

Mechanosynthesis the way forward to make a Nanofactory

A Time For AI

Sramana Mitra05.29.09, 06:00 AM EDT

Artificial intelligence, long considered obscure and unmonetizable, is starting to find its groove.

For the longest time, artificial intelligence has remained in the domain of academics, considered untouchable by investor types. Increasingly, though, this somewhat esoteric science is finding its way into crisp business applications that translate into hard dollars.

I have always considered the concept of a personal assistant--a digital servant of sorts who takes care of all my repetitive activities--extremely attractive. Recently, I met another crazy entrepreneur who has a surprisingly similar view of AI.

Patrick Grady, chief executive of Rearden Commerce, has been enchanted by the idea of a personal digital assistant ever since the early 1990s when he invested in General Magic, a company incubated inside AppleAAPL - news people ) and later spun off. General Magic pioneered the concept of a "personal intelligent communicator" with roughly the same notion of an AI-based intelligent device. That work was a precursor to what later became the PDA, and has today evolved into the smart phone.

Through the years, though, the path of entrepreneurship has been littered with failed experiments in commercializing AI. But it seems that a change is in the air.

Grady's Rearden Commerce is certainly one of the most prominent flag-bearers of AI applications that are making good on the long-time promise. Simply put, Rearden Commerce offers a platform for managing business travel and services procurement for "prosumers" (professionals who are also consumers at work) in an automated mode, the same way a personal secretary would. Book air tickets. Reserve hotel rooms. Make dinner reservations. Get event tickets. File expense reports. So on and so forth.

The company has raised $200 million in venture capital, and has partnerships with American Express ( AXP - news -people ) and Chase. "If you go back nine quarters, we had 12 customers and tens of thousands of users. Today we have well over 4,000 corporations; more than half of that is from American Express. We now have 2.5 million prosumer knowledge workers on the network, and that is before we go live with Chase," boasts Grady. But it's a boast with substance: Rearden has clearly managed to monetize his AI vision at a significant scale by bringing under one umbrella a vast network of merchants and providers and then negotiating discounts as well as enforcing policies on behalf of its client corporations.

What's the revenue model? Grady explains: "In the enterprise and higher end mid-market B2B, we charge a subscription amount based upon the amount of spend a company puts through our system and how much they save. We try to deliver roughly a 10x cash-on-cash annual return. If we are going to save you $20 million, we would like to get a couple of million dollars a year from you.

"On the back end, we receive transaction revenue from the merchants and suppliers where there is not an existing relationship," Grady continues. "If a new customer has an existing relationship with FedEx ( FDX - news people ), we do not get involved."

Rearden also aims to cater to the SME market. "When you get to the small-business market, things change. Small businesses do not have the same buying power as larger businesses. We negotiate significant discounts that you will never get on your own, pass most of that on to you and also take a piece for ourselves."

Next stop: taking the platform to consumers. "We are going live in the consumer space in a couple of months with Chase and their consumer card members. It will be the Chase brand powered by Rearden Commerce, just like we power American Express and other brands. That will be a combination of contextual advertising revenue and transaction revenue share. Partners such as American Express and Chase pay us for the right to distribute our product." (Read my interview with Patrick Grady here.)

In my January 2008 column, "Connecting You With Your Intimate Bot," I wrote:

"In a Web 3.0 world, then, a personalized travel agent will help you find and book a highly customized itinerary, leveraging all the power of previous generations of Web technology--searching (both generic and vertical), community building, content and commerce. That's how I get Web 3.0=(4C+P+VS)--the sum of content, commerce, community and context, with personalization and vertical search.

"This is complex technology, requiring sophisticated artificial-intelligence algorithms. After all, your Web 3.0 travel agent will not be a "person" but a "bot," or intelligent agent.

"But I suspect you will like your travel bot. And your career bot. And your shopping bot."

For the moment, Rearden's bot seems to be the closest to this idea, although Patrick Grady has aspirations of delivering a great deal more in terms of personalization as part of the Rearden Personal Assistant platform--a dream that is at the same time far-fetched and oddly within reach.


Sramana Mitra is a technology entrepreneur and strategy consultant in Silicon Valley. She has founded three companies and writes a business blog, Sramana Mitra on StrategyShe has a master's degree in electrical engineering and computer science from the Massachusetts Institute of Technology. Her first book, Entrepreneur Journeys (Volume One), is available from, as is her second book, Bootstrapping, Weapon Of Mass Reconstruction.

See Also:

Connecting You With Your Intimate Bot

The Next Frontier in Search Marketing

Consolidation Looms for SAAS

Plastic Logic's Touch-Screen E-Reader

Wednesday, May 27, 2009

Plastic Logic's Touch-Screen E-Reader

The company hopes to carve out a niche with its touch-based interface.

By Kate Greene

smaller text tool iconmedium text tool iconlarger text tool icon
Paper thin: Plastic Logic's e-reader is as thick as six credit cards. 
Credit: Plastic Logic
video Click here to see Plastic Logic's e-reader interface in action.

It's still early days for e-readers, and consumers can only choose between a few chunky-looking models. But by next year, Plastic Logic, based in Cambridge, U.K., will start selling a sleek e-reader that's the size of a standard sheet of paper and as thin as about six credit cards, and weighs less than a pound. The design of the device could help win over some customers, but Steven Glass, head of user experience at Plastic Logic, believes that the user interface developed for the device will play just as crucial a role.

On Wednesday, Plastic Logic will demo its new interface for the first time, at the All Things Digital D7 conference, in San Diego. The interface includes a touch screen to let users add notes to documents and save them even when the documents are transferred to another device or computer.

As with both the Kindle and the Sony Reader, Plastic Logic's display is built using E-ink: black and white microcapsules are suspended in a liquid and controlled using an electric charge. When a charge is applied, the microcapsules assume their position and form black text on a white background. However, in Plastic Logic's reader, the E-ink is deposited on a lightweight plastic backplane instead of on a glass backplane. Plastic Logic says that the plastic backplane allows for a larger reading area without adding more weight or bulk, and this makes the device more robust.

Plastic Logic hopes to further distinguish its reader from Amazon's Kindle and the Sony Reader by targeting those who read business documents created using Microsoft Office and Adobe Acrobat, as well as image files and standard e-reader files. The goal is to eliminate "the huge stack of papers that people take with them when they travel," says Glass. Many people need to sort through thousands of documents quickly, he adds, and want to mark them up by circling or underlining items or by adding notes.

In a demonstration at Plastic Logic's Mountain View, CA, facility last week, Glass showed off the upcoming reader. After the device starts up, the left side of the screen shows documents, including newspapers and books, organized in several different ways. For example, it shows the most recently transferred documents; a series of drop-down folders, similar to those of the file manager on personal computers; and a calendar with documents assigned to specific days. The right side of the screen shows icons for different documents, with their titles below. The interface also supports a search function, and when that is selected, a keyboard pops up on the screen for entering text.

When reading a document, a person "turns" the page by flicking a finger across the screen, and she can skip to a page number using a hidden toolbar that pops out of the right side of the screen when that side is touched. Depending on the size of the document, the page numbers are presented either one at a time or in groups of 10, such as 50 through 59. Tapping the screen further breaks it down to individual pages. Because marking up documents is an important feature of the interface, users can see, on these page numbers, which pages have been altered or bookmarked and can skip directly to them.

While the Plastic Logic home page interface allows for more flexibility than that of the Kindle, which offers limited ways to find documents, it appears less elegant and is slightly more cluttered.

Mary Tripsac a professor of business administration at Harvard University, who studies the e-reader industry, says that companies are still trying to figure out the best interfaces for e-readers. "I don't think anyone's gotten it right yet," she says, although she adds that it is a good idea to make it easy to keep track of annotations to documents.

At this stage, the interface may not be enough to distinguish one e-reader product from another, says Susan Kevorkian, an analyst at research firm IDC. "The primary factors for the e-book reader market have been content availability and device price," she says. Plastic Logic has so far announced publishing partnerships with several Detroit newspapers, theFinancial TimesUSA Today, and content aggregators such as Ingram Digital, LibreDigital, and Zinio. The price has not been finalized, but Glass says that it will be close to that of other e-readers on the market.

Ultimately, however, Kevorkian believes that the user interface will emerge as an important differentiator for e-readers. "Being able to comfortably annotate documents and have those changes preserved after transferring the document to another device, like a PC, boosts the utility of the reader considerably in terms of conforming to usage preferences and the larger ecosystem of a user's devices," she says.

Plastic Logic's device will be able to store four gigabytes of data and will have a Wi-Fi connection, although Glass wouldn't confirm that it would support Bluetooth or cellular wireless connectivity. It will connect to a computer via a USB wire to transfer documents and recharge. And since the device only uses electricity when it refreshes a page, it can go for days without a charge. As with the Kindle, its black and white display is able to refresh in less than a second, but if the graphics on the page are more complicated or it needs to switch from portrait to landscape, it takes a little longer.

There is plenty of room at the bottom: A talk by Richard Feynman considered by many as the father of Nanotechnology

There's Plenty of Room at the Bottom

An Invitation to Enter a New Field of Physics

 by Richard P. Feynman

This transcript of the classic talk that Richard Feynman gave on December 29th 1959 at the annual meeting of the American Physical Society at theCalifornia Institute of Technology (Caltech) was first published in the February 1960 issue of Caltech's Engineering and Science, which owns the copyright. It has been made available on the web at with their kind permission.

Information on the Feynman Prizes

Links to pages on Feynman

For an account of the talk and how people reacted to it, see chapter 4 of Nano! by Ed Regis, Little/Brown 1995. An excellent technical introduction to nanotechnology is Nanosystems: molecular machinery, manufacturing, and computation by K. Eric Drexler, Wiley 1992.

I imagine experimental physicists must often look with envy at men like Kamerlingh Onnes, who discovered a field like low temperature, which seems to be bottomless and in which one can go down and down. Such a man is then a leader and has some temporary monopoly in a scientific adventure. Percy Bridgman, in designing a way to obtain higher pressures, opened up another new field and was able to move into it and to lead us all along. The development of ever higher vacuum was a continuing development of the same kind.

I would like to describe a field, in which little has been done, but in which an enormous amount can be done in principle. This field is not quite the same as the others in that it will not tell us much of fundamental physics (in the sense of, ``What are the strange particles?'') but it is more like solid-state physics in the sense that it might tell us much of great interest about the strange phenomena that occur in complex situations. Furthermore, a point that is most important is that it would have an enormous number of technical applications.

What I want to talk about is the problem of manipulating and controlling things on a small scale.

As soon as I mention this, people tell me about miniaturization, and how far it has progressed today. They tell me about electric motors that are the size of the nail on your small finger. And there is a device on the market, they tell me, by which you can write the Lord's Prayer on the head of a pin. But that's nothing; that's the most primitive, halting step in the direction I intend to discuss. It is a staggeringly small world that is below. In the year 2000, when they look back at this age, they will wonder why it was not until the year 1960 that anybody began seriously to move in this direction.

Why cannot we write the entire 24 volumes of the Encyclopedia Brittanica on the head of a pin?

Let's see what would be involved. The head of a pin is a sixteenth of an inch across. If you magnify it by 25,000 diameters, the area of the head of the pin is then equal to the area of all the pages of the Encyclopaedia Brittanica. Therefore, all it is necessary to do is to reduce in size all the writing in the Encyclopaedia by 25,000 times. Is that possible? The resolving power of the eye is about 1/120 of an inch---that is roughly the diameter of one of the little dots on the fine half-tone reproductions in the Encyclopaedia. This, when you demagnify it by 25,000 times, is still 80 angstroms in diameter---32 atoms across, in an ordinary metal. In other words, one of those dots still would contain in its area 1,000 atoms. So, each dot can easily be adjusted in size as required by the photoengraving, and there is no question that there is enough room on the head of a pin to put all of the Encyclopaedia Brittanica.

Furthermore, it can be read if it is so written. Let's imagine that it is written in raised letters of metal; that is, where the black is in the Encyclopedia, we have raised letters of metal that are actually 1/25,000 of their ordinary size. How would we read it?

If we had something written in such a way, we could read it using techniques in common use today. (They will undoubtedly find a better way when we do actually have it written, but to make my point conservatively I shall just take techniques we know today.) We would press the metal into a plastic material and make a mold of it, then peel the plastic off very carefully, evaporate silica into the plastic to get a very thin film, then shadow it by evaporating gold at an angle against the silica so that all the little letters will appear clearly, dissolve the plastic away from the silica film, and then look through it with an electron microscope!

There is no question that if the thing were reduced by 25,000 times in the form of raised letters on the pin, it would be easy for us to read it today. Furthermore; there is no question that we would find it easy to make copies of the master; we would just need to press the same metal plate again into plastic and we would have another copy.

How do we write small?

The next question is: How do we write it? We have no standard technique to do this now. But let me argue that it is not as difficult as it first appears to be. We can reverse the lenses of the electron microscope in order to demagnify as well as magnify. A source of ions, sent through the microscope lenses in reverse, could be focused to a very small spot. We could write with that spot like we write in a TV cathode ray oscilloscope, by going across in lines, and having an adjustment which determines the amount of material which is going to be deposited as we scan in lines.

This method might be very slow because of space charge limitations. There will be more rapid methods. We could first make, perhaps by some photo process, a screen which has holes in it in the form of the letters. Then we would strike an arc behind the holes and draw metallic ions through the holes; then we could again use our system of lenses and make a small image in the form of ions, which would deposit the metal on the pin.

A simpler way might be this (though I am not sure it would work): We take light and, through an optical microscope running backwards, we focus it onto a very small photoelectric screen. Then electrons come away from the screen where the light is shining. These electrons are focused down in size by the electron microscope lenses to impinge directly upon the surface of the metal. Will such a beam etch away the metal if it is run long enough? I don't know. If it doesn't work for a metal surface, it must be possible to find some surface with which to coat the original pin so that, where the electrons bombard, a change is made which we could recognize later.

There is no intensity problem in these devices---not what you are used to in magnification, where you have to take a few electrons and spread them over a bigger and bigger screen; it is just the opposite. The light which we get from a page is concentrated onto a very small area so it is very intense. The few electrons which come from the photoelectric screen are demagnified down to a very tiny area so that, again, they are very intense. I don't know why this hasn't been done yet!

That's the Encyclopaedia Brittanica on the head of a pin, but let's consider all the books in the world. The Library of Congress has approximately 9 million volumes; the British Museum Library has 5 million volumes; there are also 5 million volumes in the National Library in France. Undoubtedly there are duplications, so let us say that there are some 24 million volumes of interest in the world.

What would happen if I print all this down at the scale we have been discussing? How much space would it take? It would take, of course, the area of about a million pinheads because, instead of there being just the 24 volumes of the Encyclopaedia, there are 24 million volumes. The million pinheads can be put in a square of a thousand pins on a side, or an area of about 3 square yards. That is to say, the silica replica with the paper-thin backing of plastic, with which we have made the copies, with all this information, is on an area of approximately the size of 35 pages of the Encyclopaedia. That is about half as many pages as there are in this magazine. All of the information which all of mankind has every recorded in books can be carried around in a pamphlet in your hand---and not written in code, but a simple reproduction of the original pictures, engravings, and everything else on a small scale without loss of resolution.

What would our librarian at Caltech say, as she runs all over from one building to another, if I tell her that, ten years from now, all of the information that she is struggling to keep track of--- 120,000 volumes, stacked from the floor to the ceiling, drawers full of cards, storage rooms full of the older books---can be kept on just one library card! When the University of Brazil, for example, finds that their library is burned, we can send them a copy of every book in our library by striking off a copy from the master plate in a few hours and mailing it in an envelope no bigger or heavier than any other ordinary air mail letter.

Now, the name of this talk is ``There is Plenty of Room at the Bottom''---not just ``There is Room at the Bottom.'' What I have demonstrated is that there is room---that you can decrease the size of things in a practical way. I now want to show that there is plenty of room. I will not now discuss how we are going to do it, but only what is possible in principle---in other words, what is possible according to the laws of physics. I am not inventing anti-gravity, which is possible someday only if the laws are not what we think. I am telling you what could be done if the laws are what we think; we are not doing it simply because we haven't yet gotten around to it.

Information on a small scale

Suppose that, instead of trying to reproduce the pictures and all the information directly in its present form, we write only the information content in a code of dots and dashes, or something like that, to represent the various letters. Each letter represents six or seven ``bits'' of information; that is, you need only about six or seven dots or dashes for each letter. Now, instead of writing everything, as I did before, on the surface of the head of a pin, I am going to use the interior of the material as well.

Let us represent a dot by a small spot of one metal, the next dash, by an adjacent spot of another metal, and so on. Suppose, to be conservative, that a bit of information is going to require a little cube of atoms 5 times 5 times 5---that is 125 atoms. Perhaps we need a hundred and some odd atoms to make sure that the information is not lost through diffusion, or through some other process.

I have estimated how many letters there are in the Encyclopaedia, and I have assumed that each of my 24 million books is as big as an Encyclopaedia volume, and have calculated, then, how many bits of information there are (10^15). For each bit I allow 100 atoms. And it turns out that all of the information that man has carefully accumulated in all the books in the world can be written in this form in a cube of material one two-hundredth of an inch wide--- which is the barest piece of dust that can be made out by the human eye. So there is plenty of room at the bottom! Don't tell me about microfilm!

This fact---that enormous amounts of information can be carried in an exceedingly small space---is, of course, well known to the biologists, and resolves the mystery which existed before we understood all this clearly, of how it could be that, in the tiniest cell, all of the information for the organization of a complex creature such as ourselves can be stored. All this information---whether we have brown eyes, or whether we think at all, or that in the embryo the jawbone should first develop with a little hole in the side so that later a nerve can grow through it---all this information is contained in a very tiny fraction of the cell in the form of long-chain DNA molecules in which approximately 50 atoms are used for one bit of information about the cell.

Better electron microscopes

If I have written in a code, with 5 times 5 times 5 atoms to a bit, the question is: How could I read it today? The electron microscope is not quite good enough, with the greatest care and effort, it can only resolve about 10 angstroms. I would like to try and impress upon you while I am talking about all of these things on a small scale, the importance of improving the electron microscope by a hundred times. It is not impossible; it is not against the laws of diffraction of the electron. The wave length of the electron in such a microscope is only 1/20 of an angstrom. So it should be possible to see the individual atoms. What good would it be to see individual atoms distinctly?

We have friends in other fields---in biology, for instance. We physicists often look at them and say, ``You know the reason you fellows are making so little progress?'' (Actually I don't know any field where they are making more rapid progress than they are in biology today.) ``You should use more mathematics, like we do.'' They could answer us---but they're polite, so I'll answer for them: ``What you should do in order for us to make more rapid progress is to make the electron microscope 100 times better.''

What are the most central and fundamental problems of biology today? They are questions like: What is the sequence of bases in the DNA? What happens when you have a mutation? How is the base order in the DNA connected to the order of amino acids in the protein? What is the structure of the RNA; is it single-chain or double-chain, and how is it related in its order of bases to the DNA? What is the organization of the microsomes? How are proteins synthesized? Where does the RNA go? How does it sit? Where do the proteins sit? Where do the amino acids go in? In photosynthesis, where is the chlorophyll; how is it arranged; where are the carotenoids involved in this thing? What is the system of the conversion of light into chemical energy?

It is very easy to answer many of these fundamental biological questions; you just look at the thing! You will see the order of bases in the chain; you will see the structure of the microsome. Unfortunately, the present microscope sees at a scale which is just a bit too crude. Make the microscope one hundred times more powerful, and many problems of biology would be made very much easier. I exaggerate, of course, but the biologists would surely be very thankful to you---and they would prefer that to the criticism that they should use more mathematics.

The theory of chemical processes today is based on theoretical physics. In this sense, physics supplies the foundation of chemistry. But chemistry also has analysis. If you have a strange substance and you want to know what it is, you go through a long and complicated process of chemical analysis. You can analyze almost anything today, so I am a little late with my idea. But if the physicists wanted to, they could also dig under the chemists in the problem of chemical analysis. It would be very easy to make an analysis of any complicated chemical substance; all one would have to do would be to look at it and see where the atoms are. The only trouble is that the electron microscope is one hundred times too poor. (Later, I would like to ask the question: Can the physicists do something about the third problem of chemistry---namely, synthesis? Is there a physical way to synthesize any chemical substance?

The reason the electron microscope is so poor is that the f- value of the lenses is only 1 part to 1,000; you don't have a big enough numerical aperture. And I know that there are theorems which prove that it is impossible, with axially symmetrical stationary field lenses, to produce an f-value any bigger than so and so; and therefore the resolving power at the present time is at its theoretical maximum. But in every theorem there are assumptions. Why must the field be symmetrical? I put this out as a challenge: Is there no way to make the electron microscope more powerful?

The marvelous biological system

The biological example of writing information on a small scale has inspired me to think of something that should be possible. Biology is not simply writing information; it is doing something about it. A biological system can be exceedingly small. Many of the cells are very tiny, but they are very active; they manufacture various substances; they walk around; they wiggle; and they do all kinds of marvelous things---all on a very small scale. Also, they store information. Consider the possibility that we too can make a thing very small which does what we want---that we can manufacture an object that maneuvers at that level!

There may even be an economic point to this business of making things very small. Let me remind you of some of the problems of computing machines. In computers we have to store an enormous amount of information. The kind of writing that I was mentioning before, in which I had everything down as a distribution of metal, is permanent. Much more interesting to a computer is a way of writing, erasing, and writing something else. (This is usually because we don't want to waste the material on which we have just written. Yet if we could write it in a very small space, it wouldn't make any difference; it could just be thrown away after it was read. It doesn't cost very much for the material).

Miniaturizing the computer

I don't know how to do this on a small scale in a practical way, but I do know that computing machines are very large; they fill rooms. Why can't we make them very small, make them of little wires, little elements---and by little, I mean little. For instance, the wires should be 10 or 100 atoms in diameter, and the circuits should be a few thousand angstroms across. Everybody who has analyzed the logical theory of computers has come to the conclusion that the possibilities of computers are very interesting---if they could be made to be more complicated by several orders of magnitude. If they had millions of times as many elements, they could make judgments. They would have time to calculate what is the best way to make the calculation that they are about to make. They could select the method of analysis which, from their experience, is better than the one that we would give to them. And in many other ways, they would have new qualitative features.

If I look at your face I immediately recognize that I have seen it before. (Actually, my friends will say I have chosen an unfortunate example here for the subject of this illustration. At least I recognize that it is a man and not an apple.) Yet there is no machine which, with that speed, can take a picture of a face and say even that it is a man; and much less that it is the same man that you showed it before---unless it is exactly the same picture. If the face is changed; if I am closer to the face; if I am further from the face; if the light changes---I recognize it anyway. Now, this little computer I carry in my head is easily able to do that. The computers that we build are not able to do that. The number of elements in this bone box of mine are enormously greater than the number of elements in our ``wonderful'' computers. But our mechanical computers are too big; the elements in this box are microscopic. I want to make some that are submicroscopic.

If we wanted to make a computer that had all these marvelous extra qualitative abilities, we would have to make it, perhaps, the size of the Pentagon. This has several disadvantages. First, it requires too much material; there may not be enough germanium in the world for all the transistors which would have to be put into this enormous thing. There is also the problem of heat generation and power consumption; TVA would be needed to run the computer. But an even more practical difficulty is that the computer would be limited to a certain speed. Because of its large size, there is finite time required to get the information from one place to another. The information cannot go any faster than the speed of light---so, ultimately, when our computers get faster and faster and more and more elaborate, we will have to make them smaller and smaller.

But there is plenty of room to make them smaller. There is nothing that I can see in the physical laws that says the computer elements cannot be made enormously smaller than they are now. In fact, there may be certain advantages.

Miniaturization by evaporation

How can we make such a device? What kind of manufacturing processes would we use? One possibility we might consider, since we have talked about writing by putting atoms down in a certain arrangement, would be to evaporate the material, then evaporate the insulator next to it. Then, for the next layer, evaporate another position of a wire, another insulator, and so on. So, you simply evaporate until you have a block of stuff which has the elements--- coils and condensers, transistors and so on---of exceedingly fine dimensions.

But I would like to discuss, just for amusement, that there are other possibilities. Why can't we manufacture these small computers somewhat like we manufacture the big ones? Why can't we drill holes, cut things, solder things, stamp things out, mold different shapes all at an infinitesimal level? What are the limitations as to how small a thing has to be before you can no longer mold it? How many times when you are working on something frustratingly tiny like your wife's wrist watch, have you said to yourself, ``If I could only train an ant to do this!'' What I would like to suggest is the possibility of training an ant to train a mite to do this. What are the possibilities of small but movable machines? They may or may not be useful, but they surely would be fun to make.

Consider any machine---for example, an automobile---and ask about the problems of making an infinitesimal machine like it. Suppose, in the particular design of the automobile, we need a certain precision of the parts; we need an accuracy, let's suppose, of 4/10,000 of an inch. If things are more inaccurate than that in the shape of the cylinder and so on, it isn't going to work very well. If I make the thing too small, I have to worry about the size of the atoms; I can't make a circle of ``balls'' so to speak, if the circle is too small. So, if I make the error, corresponding to 4/10,000 of an inch, correspond to an error of 10 atoms, it turns out that I can reduce the dimensions of an automobile 4,000 times, approximately---so that it is 1 mm. across. Obviously, if you redesign the car so that it would work with a much larger tolerance, which is not at all impossible, then you could make a much smaller device.

It is interesting to consider what the problems are in such small machines. Firstly, with parts stressed to the same degree, the forces go as the area you are reducing, so that things like weight and inertia are of relatively no importance. The strength of material, in other words, is very much greater in proportion. The stresses and expansion of the flywheel from centrifugal force, for example, would be the same proportion only if the rotational speed is increased in the same proportion as we decrease the size. On the other hand, the metals that we use have a grain structure, and this would be very annoying at small scale because the material is not homogeneous. Plastics and glass and things of this amorphous nature are very much more homogeneous, and so we would have to make our machines out of such materials.

There are problems associated with the electrical part of the system---with the copper wires and the magnetic parts. The magnetic properties on a very small scale are not the same as on a large scale; there is the ``domain'' problem involved. A big magnet made of millions of domains can only be made on a small scale with one domain. The electrical equipment won't simply be scaled down; it has to be redesigned. But I can see no reason why it can't be redesigned to work again.

Problems of lubrication

Lubrication involves some interesting points. The effective viscosity of oil would be higher and higher in proportion as we went down (and if we increase the speed as much as we can). If we don't increase the speed so much, and change from oil to kerosene or some other fluid, the problem is not so bad. But actually we may not have to lubricate at all! We have a lot of extra force. Let the bearings run dry; they won't run hot because the heat escapes away from such a small device very, very rapidly.

This rapid heat loss would prevent the gasoline from exploding, so an internal combustion engine is impossible. Other chemical reactions, liberating energy when cold, can be used. Probably an external supply of electrical power would be most convenient for such small machines.

What would be the utility of such machines? Who knows? Of course, a small automobile would only be useful for the mites to drive around in, and I suppose our Christian interests don't go that far. However, we did note the possibility of the manufacture of small elements for computers in completely automatic factories, containing lathes and other machine tools at the very small level. The small lathe would not have to be exactly like our big lathe. I leave to your imagination the improvement of the design to take full advantage of the properties of things on a small scale, and in such a way that the fully automatic aspect would be easiest to manage.

A friend of mine (Albert R. Hibbs) suggests a very interesting possibility for relatively small machines. He says that, although it is a very wild idea, it would be interesting in surgery if you could swallow the surgeon. You put the mechanical surgeon inside the blood vessel and it goes into the heart and ``looks'' around. (Of course the information has to be fed out.) It finds out which valve is the faulty one and takes a little knife and slices it out. Other small machines might be permanently incorporated in the body to assist some inadequately-functioning organ.

Now comes the interesting question: How do we make such a tiny mechanism? I leave that to you. However, let me suggest one weird possibility. You know, in the atomic energy plants they have materials and machines that they can't handle directly because they have become radioactive. To unscrew nuts and put on bolts and so on, they have a set of master and slave hands, so that by operating a set of levers here, you control the ``hands'' there, and can turn them this way and that so you can handle things quite nicely.

Most of these devices are actually made rather simply, in that there is a particular cable, like a marionette string, that goes directly from the controls to the ``hands.'' But, of course, things also have been made using servo motors, so that the connection between the one thing and the other is electrical rather than mechanical. When you turn the levers, they turn a servo motor, and it changes the electrical currents in the wires, which repositions a motor at the other end.

Now, I want to build much the same device---a master-slave system which operates electrically. But I want the slaves to be made especially carefully by modern large-scale machinists so that they are one-fourth the scale of the ``hands'' that you ordinarily maneuver. So you have a scheme by which you can do things at one- quarter scale anyway---the little servo motors with little hands play with little nuts and bolts; they drill little holes; they are four times smaller. Aha! So I manufacture a quarter-size lathe; I manufacture quarter-size tools; and I make, at the one-quarter scale, still another set of hands again relatively one-quarter size! This is one-sixteenth size, from my point of view. And after I finish doing this I wire directly from my large-scale system, through transformers perhaps, to the one-sixteenth-size servo motors. Thus I can now manipulate the one-sixteenth size hands.

Well, you get the principle from there on. It is rather a difficult program, but it is a possibility. You might say that one can go much farther in one step than from one to four. Of course, this has all to be designed very carefully and it is not necessary simply to make it like hands. If you thought of it very carefully, you could probably arrive at a much better system for doing such things.

If you work through a pantograph, even today, you can get much more than a factor of four in even one step. But you can't work directly through a pantograph which makes a smaller pantograph which then makes a smaller pantograph---because of the looseness of the holes and the irregularities of construction. The end of the pantograph wiggles with a relatively greater irregularity than the irregularity with which you move your hands. In going down this scale, I would find the end of the pantograph on the end of the pantograph on the end of the pantograph shaking so badly that it wasn't doing anything sensible at all.

At each stage, it is necessary to improve the precision of the apparatus. If, for instance, having made a small lathe with a pantograph, we find its lead screw irregular---more irregular than the large-scale one---we could lap the lead screw against breakable nuts that you can reverse in the usual way back and forth until this lead screw is, at its scale, as accurate as our original lead screws, at our scale.

We can make flats by rubbing unflat surfaces in triplicates together---in three pairs---and the flats then become flatter than the thing you started with. Thus, it is not impossible to improve precision on a small scale by the correct operations. So, when we build this stuff, it is necessary at each step to improve the accuracy of the equipment by working for awhile down there, making accurate lead screws, Johansen blocks, and all the other materials which we use in accurate machine work at the higher level. We have to stop at each level and manufacture all the stuff to go to the next level---a very long and very difficult program. Perhaps you can figure a better way than that to get down to small scale more rapidly.

Yet, after all this, you have just got one little baby lathe four thousand times smaller than usual. But we were thinking of making an enormous computer, which we were going to build by drilling holes on this lathe to make little washers for the computer. How many washers can you manufacture on this one lathe?

A hundred tiny hands

When I make my first set of slave ``hands'' at one-fourth scale, I am going to make ten sets. I make ten sets of ``hands,'' and I wire them to my original levers so they each do exactly the same thing at the same time in parallel. Now, when I am making my new devices one-quarter again as small, I let each one manufacture ten copies, so that I would have a hundred ``hands'' at the 1/16th size.

Where am I going to put the million lathes that I am going to have? Why, there is nothing to it; the volume is much less than that of even one full-scale lathe. For instance, if I made a billion little lathes, each 1/4000 of the scale of a regular lathe, there are plenty of materials and space available because in the billion little ones there is less than 2 percent of the materials in one big lathe.

It doesn't cost anything for materials, you see. So I want to build a billion tiny factories, models of each other, which are manufacturing simultaneously, drilling holes, stamping parts, and so on.

As we go down in size, there are a number of interesting problems that arise. All things do not simply scale down in proportion. There is the problem that materials stick together by the molecular (Van der Waals) attractions. It would be like this: After you have made a part and you unscrew the nut from a bolt, it isn't going to fall down because the gravity isn't appreciable; it would even be hard to get it off the bolt. It would be like those old movies of a man with his hands full of molasses, trying to get rid of a glass of water. There will be several problems of this nature that we will have to be ready to design for.

Rearranging the atoms

But I am not afraid to consider the final question as to whether, ultimately---in the great future---we can arrange the atoms the way we want; the very atoms, all the way down! What would happen if we could arrange the atoms one by one the way we want them (within reason, of course; you can't put them so that they are chemically unstable, for example).

Up to now, we have been content to dig in the ground to find minerals. We heat them and we do things on a large scale with them, and we hope to get a pure substance with just so much impurity, and so on. But we must always accept some atomic arrangement that nature gives us. We haven't got anything, say, with a ``checkerboard'' arrangement, with the impurity atoms exactly arranged 1,000 angstroms apart, or in some other particular pattern.

What could we do with layered structures with just the right layers? What would the properties of materials be if we could really arrange the atoms the way we want them? They would be very interesting to investigate theoretically. I can't see exactly what would happen, but I can hardly doubt that when we have some control of the arrangement of things on a small scale we will get an enormously greater range of possible properties that substances can have, and of different things that we can do.

Consider, for example, a piece of material in which we make little coils and condensers (or their solid state analogs) 1,000 or 10,000 angstroms in a circuit, one right next to the other, over a large area, with little antennas sticking out at the other end---a whole series of circuits. Is it possible, for example, to emit light from a whole set of antennas, like we emit radio waves from an organized set of antennas to beam the radio programs to Europe? The same thing would be to beam the light out in a definite direction with very high intensity. (Perhaps such a beam is not very useful technically or economically.)

I have thought about some of the problems of building electric circuits on a small scale, and the problem of resistance is serious. If you build a corresponding circuit on a small scale, its natural frequency goes up, since the wave length goes down as the scale; but the skin depth only decreases with the square root of the scale ratio, and so resistive problems are of increasing difficulty. Possibly we can beat resistance through the use of superconductivity if the frequency is not too high, or by other tricks.

Atoms in a small world

When we get to the very, very small world---say circuits of seven atoms---we have a lot of new things that would happen that represent completely new opportunities for design. Atoms on a small scale behave like nothing on a large scale, for they satisfy the laws of quantum mechanics. So, as we go down and fiddle around with the atoms down there, we are working with different laws, and we can expect to do different things. We can manufacture in different ways. We can use, not just circuits, but some system involving the quantized energy levels, or the interactions of quantized spins, etc.

Another thing we will notice is that, if we go down far enough, all of our devices can be mass produced so that they are absolutely perfect copies of one another. We cannot build two large machines so that the dimensions are exactly the same. But if your machine is only 100 atoms high, you only have to get it correct to one-half of one percent to make sure the other machine is exactly the same size---namely, 100 atoms high!

At the atomic level, we have new kinds of forces and new kinds of possibilities, new kinds of effects. The problems of manufacture and reproduction of materials will be quite different. I am, as I said, inspired by the biological phenomena in which chemical forces are used in repetitious fashion to produce all kinds of weird effects (one of which is the author).

The principles of physics, as far as I can see, do not speak against the possibility of maneuvering things atom by atom. It is not an attempt to violate any laws; it is something, in principle, that can be done; but in practice, it has not been done because we are too big.

Ultimately, we can do chemical synthesis. A chemist comes to us and says, ``Look, I want a molecule that has the atoms arranged thus and so; make me that molecule.'' The chemist does a mysterious thing when he wants to make a molecule. He sees that it has got that ring, so he mixes this and that, and he shakes it, and he fiddles around. And, at the end of a difficult process, he usually does succeed in synthesizing what he wants. By the time I get my devices working, so that we can do it by physics, he will have figured out how to synthesize absolutely anything, so that this will really be useless.

But it is interesting that it would be, in principle, possible (I think) for a physicist to synthesize any chemical substance that the chemist writes down. Give the orders and the physicist synthesizes it. How? Put the atoms down where the chemist says, and so you make the substance. The problems of chemistry and biology can be greatly helped if our ability to see what we are doing, and to do things on an atomic level, is ultimately developed---a development which I think cannot be avoided.

Now, you might say, ``Who should do this and why should they do it?'' Well, I pointed out a few of the economic applications, but I know that the reason that you would do it might be just for fun. But have some fun! Let's have a competition between laboratories. Let one laboratory make a tiny motor which it sends to another lab which sends it back with a thing that fits inside the shaft of the first motor.

High school competition

Just for the fun of it, and in order to get kids interested in this field, I would propose that someone who has some contact with the high schools think of making some kind of high school competition. After all, we haven't even started in this field, and even the kids can write smaller than has ever been written before. They could have competition in high schools. The Los Angeles high school could send a pin to the Venice high school on which it says, ``How's this?'' They get the pin back, and in the dot of the ``i'' it says, ``Not so hot.''

Perhaps this doesn't excite you to do it, and only economics will do so. Then I want to do something; but I can't do it at the present moment, because I haven't prepared the ground. It is my intention to offer a prize of $1,000 to the first guy who can take the information on the page of a book and put it on an area 1/25,000 smaller in linear scale in such manner that it can be read by an electron microscope.

And I want to offer another prize---if I can figure out how to phrase it so that I don't get into a mess of arguments about definitions---of another $1,000 to the first guy who makes an operating electric motor---a rotating electric motor which can be controlled from the outside and, not counting the lead-in wires, is only 1/64 inch cube.

I do not expect that such prizes will have to wait very long for claimants.