Underflow | 7
In this episode: Bitcoin inventor Toomas Peters has finished his PhD thesis and has launched his new side-project, called Cicada_3301. He is wondering what to do next when he receives an email from his old friend Lars Lipp proposing that they work together on a new protocol allowing large sums of money to be transferred instantly between international banks on behalf of Russian oligarchs.
March 2012 — Princeton, NJ, and Tallinn, Estonia
AFTER A SPIKE to thirty dollars at the end of 2011, the price of Bitcoin was hovering around five dollars in early 2012, making Toomas Peters's holding of the first million Bitcoins worth around five million dollars; which was not a bad return on an asset whose value had been zero when Peters launched Bitcoin just three years earlier.
True, daily trading volumes were still so small that the price would probably crash right back down to zero if Peters tried to liquidate his million Bitcoins all at once; nor could his capital gains from Bitcoin compare with his capital gains from Lynxite, the Internet platform which he had coded in 2003 and sold for half a billion dollars in 2008. But he sensed that the traction was there, and perhaps the critical mass was almost there, for Bitcoin to grow by further orders of magnitude — when the stars were in alignment. He would hold on to his Bitcoins for the time being.
His PhD thesis was, thought Peters, ready for submission, although he had been working seriously on it for less than a year. In it, he discussed what might happen if an instance of artificial intelligence was instructed to rewrite its own source-code. He had sold an early draft of the thesis to a start-up in Silicon Valley called NepoAI, which had subsequently invited him to join its advisory board; and just lately he had been using NepoAI's algorithms and computing power to generate complex logical puzzles which he had been posting to the Internet with the eventual aim of creating a sort of LinkedIn for people with IQs of 150-plus. For this project he was using the pseudonym Cicada_3301, which had once been the home telephone number of Jorge Luis Borges.
But still, Peters needed a next big thing; and possibly here it was, he thought, as he read an email newly arrived in his inbox from his one-time personal assistant, Lars Lipp. The two kept up with one another, so Peters knew that Lipp had taken a job with Bank Livonia after the failure of Peter's AI company, EstonAI, and also that Lipp had left the bank just a few months ago to run a start-up in Tallinn for an unnamed Russian billionaire.
Now, apparently, Lipp had a business proposition. His email read as follows:
T — I need your genius for speed and scale. My principals will be moving sums of $10-100 million through the European banking system. Seems to me such transfers could be done in seconds, but in practice banks can take hours or even days to process payments owing to manual verification. Could you/we design an online interface or app enabling the sender of funds (rather than the bank) to create and encode the data needed to route a transfer through one or a succession of banks to the designated recipient, such that the bank(s) will execute the transfer automatically and instantaneously on receipt of the sender's instruction, without human interference and resulting delay? We can trial the protocol with Bank Livonia, then generalise it to all banks — L
Peters thought for a moment. He would need to know much more about how Swift and other clearing systems worked. But if one conceptualised a bank transfer as a sort of one-time cryptocurrency, then the problem was readily soluble in principle.
Every bank transfer took the form of an encoded message containing the identity of the sender, an instruction to debit the stated amount from the sender's account, a reference number for the transaction, the identity of the recipient, and the number of the recipient's account.
There was no logical reason why such a message should not be created by the sender rather than the bank. The crucial elements were that the sender should not be able to specify any account other than their own from which the debit should be made; that the message should be securely encrypted; and that the sender must assume the entire transaction-risk as the price of bypassing the bank's controls.
Peters's reply to Lipp went as follows:
L — Yes. You catch me at a good time. I've pretty much got the thesis done, I'd like to get back to Estonia for a while, and I'd be interested in doing something that could scale across the whole banking sector. How about I block in three months, starting June 1, and you find me a neighbouring office, and we see how this goes — T
It was, in fact, Pavel Gudichev, Lipp's boss, who had raised the question of speeding up bank transfers, and Lipp had grasped his point at once. Already, to test their systems, they had sold two cargoes of Russian cement to a Serbian buyer, and, each time, the payments had taken two days to clear in each direction.
In Gudichev's view, if you could contract to sell ten thousand tons of cement instantly, why could you not receive payment instantly? "We ourselves do the risk assessment", Gudichev had said. "We trust the supplier to deliver and we trust the buyer to pay. What does the bank contribute by sitting on our money? How can we improve on this?"
When Gudichev was next in Tallinn, two days after Lipp's email exchange with Peters, Lipp explained his planned solution as follows:
"I have shared your question with my former colleague, Toomas Peters. He is also interested in the question, and he has the skills to address it. He is willing to commit three months of his time, starting in June, to creating a protocol which will allow corporate customers such as ourselves to specify and encode transfer requests in such a way that transfers can be executed automatically and instantaneously by participating banks. I myself will make a presentation to Bank Livonia suggesting that by adopting this innovation they will obtain a significant advantage over rival banks, first in Estonia and then in the whole of Europe, and I will propose that Talcow Capital Partners act as first user, test user, for the protocol. We will have to assume the risks of failure in our own transfers, and we will have to indemnify the bank against any associated risks or costs."
Gudichev said, simply: "Do it".
Lipp was surprised that Gudichev asked no questions. If their roles had been reversed Lipp would have been asking for more information about Peters, for the scope of any possible indemnity, for a schematic account of how the protocol would work, and for the likelihood that Bank Livonia would agree to give it a try.
Gudichev's apparent equanimity suggested either that he knew the answers already, or that he was indifferent to the risks, or simply that he trusted Lipp. Perhaps it was a mixture of these things.
Later that same evening Gudichev made a telephone call to a particular person in St Petersburg, using a particular telephone which Gudichev carried for such moments. The call was brief, a few words on either side:
Gudichev: The boy is doing well. We are where we want to be. We have opened the bank accounts. We have made the first trades. When I told the boy that we needed to find a way of moving our money without the banks' interfering, he understood the logic. He has a plan to do this using his friend Peters, just as you foresaw. This will happen in June. It will take three months, he says.
The Person: All good. Peters is the one we need. You have managed the boy perfectly. He is behaving well. No indiscretions so far. He may have a future. The Big House will be pleased.
If Lipp had overheard this brief telephone call, what might he have thought?
He would have bristled, certainly, at being called "the boy".
He would have been shocked to discover that Peters was in some way known to Gudichev's interlocutor, and that he himself had been recruited, at least to some extent, as a means of recruiting Peters. He would have wondered what the "Big House" was, though he might have guessed.
But even then he could have reconciled these revelations with his own understanding of what was really going on — namely, that he had been hired to assist Gudichev and a network of oligarchs in exporting billions of dollars' worth of commodities from Russia while creaming off huge commissions and evading taxes.
The opportunity to work with Peters in devising a protocol that might radically improve the efficiency of the European banking system now appeared to Lipp as an unforeseen bonus which might well, if it could be commercialised, profit him and Peters still more. The notion that they might be assisting in something altogether more sinister had not yet occurred to him.
To be continued ...
THE BROWSER: Caroline Crampton, Editor-In-Chief; Kaamya Sharma, Editor; Jodi Ettenberg, Editor-At-Large; Dan Feyer, Crossword Editor; Uri Bram, CEO & Publisher; Sylvia Bishop, Assistant Publisher; Al Breach, Founding Director; Robert Cottrell, Founding Editor.
Editorial comments and letters to the editor: editor@thebrowser.com
Technical issues and support requests: support@thebrowser.com
Or write at any time to the publisher: uri@thebrowser.com
Elsewhere on The Browser, and of possible interest to Browser subscribers: Letters To The Editor, where you will find constructive comment from fellow-subscribers; The Reader, our commonplace book of clippings and quotations; Notes, our occasional blog. You can always Give The Browser, surely the finest possible gift for discerning friends and family.