Satoshi’s Memoir Part 1
I’ve always been better with code than words. I felt now was a safe time to start clearing up some of the speculation around Bitcoin and how it came to be.
The phone rings. There was no way of knowing this would be the catalyst which would lead to the creation of Bitcoin. It was 1994 and my life was about to change forever.
I picked up the phone, my father is on the other line asking if I called 911. He said there is a police officer at the house saying someone called 911. I told him I didn’t and I’ve been on the internet all day and playing Doom. He said the police officer wants everyone in the house to come outside to ask about the 911 call and make sure everyone is safe. I found this very strange, why didn’t they just call back if someone called from the house?
I try to wake up my sister to get her downstairs, my attempts were futile. It’s the middle of the afternoon in the summer, why would she be awake.
I grab my full trashcan sitting next to my computer desk and bring it with me. My dad and the police officer are in the two-car garage with one door is open. When I enter the garage, I dump my small trashcan full of floppy disks into the trash compactor and turn it on.
The police officer proceeds to ask me if I called 911 and if there is anyone else in the house. Annoyed, I tell her, “No, it’s just me, my sister is there but, she won’t come. I’ve been on the phone all day so no one could have called.”
Before I could even take my next breath, a flood of officers pour into the garage, the first one holding up a piece of paper and yelling they have a search warrant. Shocked and confused, my father and I look at each other with puzzled expressions and worry on our faces.
I offer to call my sister on our other phone line, but one of the agents now filling the garage tells me not to touch anything. They say they are worried I may trip a switch to erase my hard drives.
Standing there, taking everything in, I faint.
When I wake up, the police and agents tell me why they are there. A group of hackers broke in the library of our metro-area university and attempted to delete the data on their servers. I was very confused. I wasn’t part of any hacking group, nor had I done anything malicious. Eventually, I found out the number my friend gave me (he found it in the phone book), which I used to access the internet with no password protection was also being used by this hacker group. When the feds did the trace on all incoming lines, one of them came back to our house… and me.
This incident shaped the rest of my life. While I awaited for my trial to run its course, I was able to borrow an amber dumb terminal and a 300 baud modem from my friend. I went on a search for the people who had done what I was accused of, and due to the amount of people online in the mid-90s, it didn’t take long to find them.
I had instant street cred from this incident, which I didn’t deserve, and subsequently was invited the join a hacking group affiliated with the other group. We would chat on BBSes, IRC and Usenet. Seeking out how to use and abuse phone systems and cyberspace. It was a rough and tumble place, but it was where I belonged. I learned to hack System 75 PBXs and setup lines we could call in to, punch in a code and it would drop to an outside line. We could call anywhere. We would set up AT&T alliance teleconference party-line calls, we’d call BBSs all over the world for free and expand our knowledge and reach. It was amazing. And then one day, I discovered the hacker manifesto in an old issue of Phrack Magazine (
http://phrack.org/issues/7/3.html). I immediately identified with it and as a result of my arrest, I took my privacy online very seriously and using many monikers over the years. I also aligned with EFF and later Freenet.
It was around this time I learned to program, first, in a scripting language for a modem software called Telix. I would write scripts to wardial entire exchanges and when computer systems answered, my scripts would try to automatically break in… Looking for PBXs we could hack into. I quickly hit limitations of the software and looked for ways to build my own software. A friend of mine told me about a programming language called C++. I was able to convince my dad to get me the Borland C++ IDE for MS-DOS and a book on C++. From that moment on, I was a computer programmer, every day, all day. I would spend 8-14 hours a day learning to code, I wrote my own modem software, a few games, and several coding demos. Demos were small programs that pushed the computer hardware of the time to achieve cool effects and show off coding skills. Two of my favorites I created were first, a flame program I called Sol, which generated flames on the screen, the second was a 3D terrain fly-over. It used a Perlin noise generator, converted it to 3D with the terrain represented as dots.
Programming lead me to a career as a developer in my senior year of high school. After high school I skipped college and went directly into my career as a software engineer mostly working with email systems. From time to time I continued to work on side projects. The most notable was my attempt at solving the RSA Factoring Challenge.
In 2004, I discovered the RSA Factoring Challenge. It was a challenge wherein an encrypted phrase was provided by the RSA Labs (the company who created the RSA asymmetric public-key cryptosystem). They had multiple challenges of varying complexity with rewards for solving the challenge, decrypting the answer and returning the encrypted strings. Essentially, what you had to do was find two near equal length prime numbers, multiply those together to get the semi-prime number provided by RSA Labs. If you were able to get the two primes, you’d be able to use them as the keys to decrypt the message.
I came up with what I think was a novel approach to a prime sieve for factoring these primes and solving the solution in a distributed, yet centralized, manner. I figured out a way to factor large primes by factoring the semi-prime from right-to-left. This method would generate a tree for factoring the number. (My first exposure to trees was the binary space partitioning method used in Doom) While factoring, you’d be able to check if a leaf on the tree was valid because all the digits (except the first) would match the semi-prime if you multiplied the two factors together. Running this on a single computer at the time for the 200 decimal digits RSA number would have taken 4 billion years. Obviously, this wouldn’t work, but I had an idea. What if I used a distributed computing approach following the SETI @ Home distributed model, where participants could run my program on computers (nodes) all over the world to solve the problem and share in the reward? People were running SETI @ Home based on an altruistic libertarian viewpoint, so the chance of making a little money seemed like it would work as an incentive for adoption and participation beyond those who simply wanted to better humanity as a reward. I also had a small network of machines I could test with too.
In 2007 the RSA Challenge was cancelled and I still hadn’t finished a few key aspects of my program:
1. C++ Rewrite: I wrote the program in Java to be platform independent. But found it slower to develop and less optimised than if I had just used C++
2. I was using a centralized database every node would connect to and update.
3. How to distribute rewards? If someone was factoring on a really fast computer 24/7, while someone else did it for an hour a day on a slow computer wouldn’t be fair to receive equal rewards once the challenge number was factored.
4. Each node would pick up a branch of numbers to process. But if they left the network, the branch wouldn’t finish processing. These branches would also never be collapsed and removed.
5. The nodes on the network may not be honourable and couldn’t be trusted to not attack the system.
6. User’s had to register using their email address, which lead to privacy concerns due to PII exposure and lack of anonymity.
Once the reward for the challenge was gone, I didn’t see the point in continuing with this project and moved on. The project I focused on next was a marketplace to sell MP3s to DJs. As I was researching for it, I fell down a rabbit hole concerning how to collect payment from customers and how to compensate the artists and labels. It was then I realized the risk was if a customer reversed the charge after paying, there was nothing to be done as they already owned the digital asset. There had to be a better way of sending money online than the current payment gateways and e-gold solutions. I set out to find a system where the payments were sufficiently irreversible and negotiated between two parties and there wasn’t a third-party required for mediation.
It wasn’t very long after the RSA Challenge ended when in early/mid 2007 I had an idea how to rework my prime sieve to solve these problems and to generate electronic cash rewards without the RSA challenge prize money. In the new system participants in the network managed the network, executed checks and balances to prevent against potential abuses of power and the reward for all this was a digital coin. Not minted by a central authority or any third party, but rather, the system itself. The first and biggest improvements I made was using IRC to allow the nodes to discover each other and communicate, the syncing of the database across all nodes and finally, finding hashcash as a system for proof-of-work when I was figuring out ways to make a version of the RSA challenge math problem. One to be completed in a predictable about of time with definable complexity.
This system was more complicated than the prime sieve as it would be a self-sustaining distributed database synchronised with a proof-of-work system. I worked on major milestones, researching each problem and coding up proofs of concept when I could isolate them. In the summer of 2008 I started compiling all my notes on the mostly working and refined code I wrote over the past year-and-a-half and wrote up a whitepaper to share my ideas in an easily digestible form with the community for feedback. I decided to keep myself as anonymous as possible to keep the project completely independent and free from entities who could be subpoenaed or held accountable in some way. It also allowed me to separate myself once network effect of Metcalfe's law took over and a sufficient community behind it.
I now turned my focus towards a catchy name for my new program and of the dozen or so names, E-Cash, Digicash, and Netcoin were the finalist. Turns out Digicash was already taken. I liked E-Cash due to my extensive background in email development, but ultimately felt “e” names were dated. So, I picked Netcoin to be the name. The next day, while working on some bit shifting code and thinking about the coin’s name, it came to me. Bitcoin.
Satoshi
