InterviewThe scientist who almost made Mark Zuckerberg president of the World Bank
The scientist who almost made Mark Zuckerberg president of the World Bank
Project Diem, the megalomaniac experiment led by Facebook, aimed to create a unified digital currency that will allow easy transfer of funds and payments between billions of people. The person who led the development of the currency was Prof. Dahlia Malkhi, one of the world's leading computer scientists. Now, in a rare interview from the bowels of the powerful technology giant, she reveals how the system was going to work, and why the fear of Facebook's power led to the collapse of the venture
Prof. Dahlia Malkhi was about to pop the champagne. The veteran computer scientist and chief technology officer of the Diem Association, Facebook's world-wide crypto project, was prepared for its grand launch. Even the giant companies that were partners in the founding of Diem, including Uber, Spotify, Shopify, and Coinbase, were ready in their own ways to launch the new payment network, which was going to reach the fingertips of two billion potential customers.
If everything went as planned, any Facebook, Messenger or WhatsApp user could download a digital wallet called Novi to their smartphone, and use it to send quickly and at zero cost the new currency Diem, whose value would be pegged to the U.S. dollar. New and private internet money, which moves within a closed, mini-sovereign financial loop, and changes hands without being affected by borders, nationality or banks. Direct competition to all payment and money transfer systems as we know them. The Diem Association not only built all the technology to enable the free movement, but also the systems that will manage and ensure the stability of the system, supervise the registration and verification of the transactions in it, issue this new money and manage the huge reserve that will back it up.
"On July 2, 2021, around 5 in the afternoon, the evening before the launch, we received a call from the U.S. Treasury Department to discuss our project," Malkhi says in an exclusive interview with Calcalist. "During the conversation, we were told that the American government does not yet have a clear framework to oversee the project, so they strongly advise us not to launch."
So, they just said that in a direct manner?
"It was heartbreaking and traumatic, both the content and the way in which it was done - in the 90th minute, after a series of very positive conversations. We specifically asked if there was anything that the project did not address, and there was not. Our leadership asked to meet with Treasury Secretary Janet Yellen and Federal Reserve Chair Jerome Powell, whom we contacted in writing and did not receive a reply. Behind the scenes we were told by internal sources that there was an order not to contact our project, not to answer any inquiries."
Thus, with one phone call, ended one of the most talked about projects in the world. Last February, Diem was dismantled and sold. In April, Malkhi, like many of her colleagues, moved on to another crypto venture as the lead researcher at Chainlink Labs. "The team that worked on the project founded various startups based on the technology that was developed," she says proudly with a touch of sorrow. "We are continuing parts of the vision, but no one is continuing the whole vision."
Who believes Zuckerberg?
Even if Malkhi finds it difficult to admit it, during the four years of the existence of the megalomaniac Diem project, the odds were stacked against it. The idea that a tech giant with a dubious reputation for user privacy would set up an international payment system with its own unique currency did not go down well with any regulator. The very pretension was seen in the media as audacity, not to mention “chutzpah”.
Suspicion did not disappear even though Diem Association was founded by Facebook as a non-profit institution, with 28 companies in founding status as equal partners, and Facebook even promised that after the launch of the system, it would step down from any leadership role in it. The fear was that the system and currency would be more popular for making payments and money transfers than any other national currency, which would give Facebook access to the data and financial activity of billions of people, challenge the dollar, weaken the power of banks, credit and payment companies, and become a focus for money laundering.
I admit that I, too, when I heard about the Facebook project, felt great discomfort.
"I did not experience Facebook. We were in an isolated building and we operated as an isolated unit. During a certain period we even operated in secrecy and it was not possible for Facebook employees to enter the building at all. I admit and confess that we felt disconnected, I did not even feel like I was joining Facebook, I joined Libra (Diem's former name). But unfortunately, the authorities did not allow us to exclude Facebook from the discussion."
Let's talk about the essence. Why do you need such a system?
"This is really the most misunderstood thing by many. The digital experience available today through the credit companies is quite good and convincing, so they did not understand why a new payment network was needed. On the other hand, the people in the Bitcoin and Blockchain worlds only used these systems for speculation and created chaos. We bridged between the two worlds."
So what does this mean?
"In systems based on Visa or MasterCard, the first thing you need is credit, and here you have already removed about 2 billion people from the system who will never have access to credit. Secondly, even those who have a privilege and a credit card, using it is long and cumbersome and the fees can reach 7%. And this is a slow process, no one has access to this information.
"With a currency like Diem, on the other hand, you take the money you have and replace it with its digital representation. From that moment on, you can send the coin to anyone in the ecosystem, without the need for approval from anyone. It is also faster, more inclusive, and its transfer fees were something like two hundredths of a percent."
For users, Diem's system offered simplicity first. Anyone who downloaded the wallet - no matter where they are in the world - can use it to buy a dollar worth of Diem coins, and immediately transfer and receive Diem coins with almost no money transfer or currency exchange cost. If the currency was widely adopted it could be the main currency of the internet. The Esperanto of money. The project's connection to payment companies, credit cards and retail companies also hinted at long-term goals: to make it a tool for paying online for services all over the world. After all, why pay Uber on a trip abroad with a foreign credit card for all the fees involved, if you can pay using the same currency and wallet all over the world.
The Byzantine Generals Problem
It is no coincidence that Malkhi headed the technological development team of the huge project. She is one of the leading and influential researchers in the world of distributed systems. After completing her three degrees in computer science at the Hebrew University of Jerusalem and teaching there, she went to Silicon Valley in the early 2000s. Throughout her extensive career, she has published more than 200 papers on distributed computing, worked at organizations such as Microsoft Research, Bell Labs and VMware Research, and in 2021 she was awarded the title of Fellow of the ACM (Association for Computing Machinery), the highest honor given by the most important association in the field of computing and computer science.
A short pause is required to understand Malkhi's field of activity: the world of computing and distributed systems sounds foreign, but it surrounds our lives. It’s the Internet and the cloud. In the age of information, developments in this field are what allow us the simultaneous, safe and synchronized experience, where a person updates a Word file in the cloud in Israel and their colleague in Ireland immediately sees the changes and adds their own. Malkhi researches and develops the ability to maintain reliability and reproduce information in such systems. For example, making sure that enough copies of our email are kept, so that if a copy is destroyed, there will be an exact copy of it somewhere else - and still one that only we will have access to.
In 2004, Malkhi joined Microsoft Laboratories and began working with Leslie Lamport, a Turing Award winner and one of the key figures in the world of distributed computing. The fruitful partnership between the two produced several developments that became the foundations and literally the gold standard in cloud computing. Their work can be viewed as dealing with one fundamental theoretical problem, which Lamport himself formulated in 1982 and which is deeply intertwined in the field of distributed computing: "the problem of the Byzantine generals."
Imagine divisions of a Byzantine army occupying a city, with each division commanded by a general, who could communicate with others only through messengers. The generals are watching over the city and have to decide on a common course of action, when only an agreed action will lead to victory and any other action will result in a catastrophic defeat. Among the generals there can be traitors who try to prevent the faithful from reaching an agreement, and therefore may sabotage the work of the messengers. In the metaphor, the generals are computers and the messengers are the internet. The messages that are sent will be the effort to convey the same message to everyone.
For years many other researchers have worked to outline the principles for the operation of distributed systems. Most of the systems born are based on a combination of decentralization and centralization: the network may be deployed everywhere, but there are still significant players who concentrate specific responsibility for reliability and the protection of information against attacks by bad actors or malfunctions. "Reliability and safety are two steps that solve the same problem," Malkhi explains. "They try to prevent malfunctions that may happen in advance, by confirming who is participating (authentication), signing information, and sometimes keeping it confidential and encrypting it. But everything can't always be prevented, sometimes there is intrusion by those who have passed all the approvals, and sometimes unexpected malfunctions just happen."
Three exciting things
Then, in 2008, Bitcoin arrived, and a decentralized payment system introduced a new technology: blockchain. This system seeks to enable the transfer of digital economic value, and to make sure that it will not be possible to "use" the same digital currency twice in a distributed system, where it is not known who and how many participants there are at any given moment, a mechanism was built that incentivizes good behavior. All information is transparent to all and available to all participants. Participation requires an investment of computing power and is rewarded through the issuance of a unique cryptocurrency - Bitcoin.
How did the research community react to the new development?
"In general, we pretty much ignored Bitcoin, it seemed to us to be something speculative, a sort of game. But in 2014, Marc Andreessen (venture capital investor and Netscape developer) published in The New York Times the article Why Bitcoin Matters, in which he claimed that Bitcoin solved the problem of Byzantine generals , and we could no longer ignore it."
And did it solve the problem?
"Our initial reaction as experts was 'no'. But then we dived deeper and understood what Bitcoin solved and what it didn't solve. Today we understand its shortcomings, for example that it is a system that wastes energy, that the race of the participants to provide a service is very expensive, and that it must clearly be slow because if it's too fast there will be collisions, but in the end I'm less interested in driving participation through tokens, and more interested in the possibilities that the technology opens up.
"From 2014, we began to look intensively at the algorithmic foundations of Bitcoin and Ethereum (another blockchain network developed in 2015), to improve them, to bridge them and also to see what parts of the vision in the field can be brought to other systems. I published a series of articles that took original and good ideas from Bitcoin and brought them to places where they can be analyzed and understood, and finally I published a work called HotStuff, which is actually an algorithm that makes it possible to build a blockchain that developers can understand, not just the experts."
This groundbreaking work caught the eye of David Marcus, formerly president of PayPal and Messenger and later head of Facebook's crypto project. In 2018, he approached Malkhi and invited her to work with them. Her HotStuff became the basis for the entire Diem project.
"Three components excited me about Marcus' invitation," says Malkhi. "The first was his vision of 'financial inclusion,' meaning technology that can reach every person through a small device. The second was the involvement of Facebook, which in retrospect killed the project but what it meant was that one of the technology giants was putting its enormous resources towards the success of the project. And the third element was The Diem consortium, which included the strongest companies from the worlds of investments, social impact, fintech and online services. This is a group that was created to open a discussion with authorities, governments and central banks all over the world and enable the launch of the platform we built."
What did you try to build?
"We tried to build in the private sector a system that produces digital money that is backed by a central bank, but not directly because we were not a bank. A system whose infrastructure is open, encourages the participation of diverse digital wallets and enables direct access and exchange of value between them. Unlike PayPal, for example, which allows the use of digital money only from one PayPal account to another.
"After all, becoming a bank is a long and expensive process that few complete. Also, more than half of the adult population in the world does not have access to financial systems. Even when they do, it is slow, it is not transparent and it is expensive, certainly when it comes to transferring money between countries or exchanging currency.
"We wanted to take deposits from people, but not for the purpose of loans or investment, but only for the purpose of transfers. We planned to manage these transfers in an open manner, to encourage participation and competition; transparent, so that anyone can see transfers between organizations; and programmable, so that everyone has access to the infrastructure and can develop ideas. It's not Bitcoin, and it's also not to bypass Fiat currencies (a state-backed currency) or license governments to print money. We just wanted to create open digital money, which would allow digital wallets to access the clearing house layer directly."
Like bitcoin, but safe
There is a fundamental difference between the Diem that Malkhi asked to build and Bitcoin. Bitcoin is a "permissionless" system, meaning one that is maintained by anyone who wants to maintain it, secured by cryptography and incentivized by "mining" Bitcoin. No one can exclude anyone from the game, there is no censorship and it is very difficult to damage the system.
In Diem, on the other hand, there is authorization, that is, defined entities that were supposed to manage and maintain the servers. If it is Visa, MasterCard, PayPal, Uber or Lyft, each of the founding companies was supposed to be responsible for the system and oversee the various information nodes. All they needed to join was to invest $10 million in purchasing Diem coins and meet certain benchmarks such as a market cap of at least $1 billion or at least 20 million customers. Facebook itself was the driving force, with the original digital wallet.
The public perception was that the project wanted to replace the existing state financial system.
"There were several reasons for this. One of them is that in June 2019 Facebook issued a statement to the world: We Re-Invent Money. It was an unbelievable mistake, to talk about reinventing money and declare it in front of the American government before you received feedback on what you are trying to do."
You didn't want to reinvent money?
"No. We wanted to launch a payments platform that enables innovation and efficiency that the private sector can understand, in contrast to the banking world which is very exclusive and uses less efficient and archaic technology."
Those who were not part of the founding members were banks and other technology giants such as Amazon, Google or Apple - and not by chance. Facebook created a new financial model, legally, organizationally and technologically complex, designed to allow it to issue a new currency, accept deposits, manage money and process transactions without being registered as a bank. This bothered many.
In June 2019, when the project became visible and its document of intent was published, Facebook received a barrage of criticism, accompanied by concern bordering on panic. The new development managed to unite against it politicians from the left and the right, regulators from all over the world and privacy experts. Everyone thought it was a dangerous idea to put the system in Facebook's hands.
From here began what, in hindsight, turned out to be a liquidation process for the project. In July 2019, immediately after the announcement, the head of the project Marcus was summoned to a hearing before the Banking Committee in the Senate. Shortly after, in October, Facebook founder Mark Zuckerberg was also summoned to another hearing before the Finance Committee in Congress. Between these two hearings, a series of regulators, including the Bank for International Settlements (BIS), the Financial Stability Board (an international body that monitors the global financial system), the French and German finance ministers and the UK's Financial Conduct Authority expressed dismay at the project.
The first effective blow was the Federal Reserve's warning that any financial institution planning to be involved in project Diem should expect stricter scrutiny of its payment activities. A short time later, PayPal, Visa, Mastercard, Stripe and eBay withdrew from the project.
What did you think when the first companies backed out?
"In internal discussions, the leadership said it was disappointing and sad, but on the other hand, it is better that organizations that are not 100 percent aligned with our goal take a step back, than to continue developing with organizations that are not completely coordinated."
In retrospect it turned out that the companies that withdrew were actually coordinated, only with the other participants. At the end of October, Zuckerberg appeared before a congressional committee, where a firm position was presented to him: the project was seen as a disturbing effort to accumulate power and should be stopped. "Given the size of the company and its reach, it should be clear why we have serious concerns about your plans to create a global digital currency that will challenge the U.S. dollar," Committee Chair Maxine Waters told Zuckerberg, adding: "You have opened a serious discussion about whether Facebook should be broken up." Congressman Brad Sherman thought that the claims of financial inclusion were cynical to say the least: "The richest man in the world comes here and hides behind the poorest people in the world and claims that he is trying to help them," he slammed Zuckerberg. "You are helping those for whom the dollar is a problematic currency - drug dealers, terrorists and those who avoid paying tax."
In May 2020, the European Central Bank published an analysis according to which, if the project is successful, Facebook could find itself managing assets worth approximately $3 trillion and reserves totaling $150 billion. Since these funds were supposed to be invested in high-quality liquid assets such as government bonds, the implication was that the organization might become one of the largest financial funds in the Eurozone and destabilize the bloc's monetary stability.
Who is bothered by 2 billion wallets
Facebook and Zuckerberg did not give up and tried to restore the project. In December 2020, they changed its name from Libra to Diem, to demonstrate distance from Facebook and greater independence. "In the first step, we tried to launch the payment system as a Swiss system, and entered into a dialogue with FINMA, a Swiss state authority that provides licenses for payment systems," Malkhi recalls. "The dialogue lasted almost a year, we introduced the payment system and made changes and improvements as required. We built all the risk management and customer protection systems so that we could launch. In March 2021, we were told that we answered all the questions and requirements and that the authorization process for the system is underway."
It did not happen.
"On March 19", Malkhi remembers each date, "towards the end of the process with the Swiss, they received a response from the United States government, which said that more time was needed to approve its part, because the payment system is expected to operate in U.S. dollars. This was a death blow to the launch in Switzerland. Such a delay without a final date in an industry that moves so fast is something that cannot be absorbed."
Did you give up?
"In coordination with the American government, we moved all operations to the United States. We teamed up with a federal bank called Silvergate to launch a payment system. This time the digitization of money was supposed to be something that the bank would do, while we would only run the transfer system jointly with them. It took us three months to readjust the system that was built for launch in Switzerland, and in coordination with the government, we announced that we expect to launch the system at the end of June 2021. The date was getting closer, and not only did we receive no objections, we received clear signals that everything was done as expected."
It's a bit like burying your head in the sand, isn't it? The public uproar did not subside.
"No, we developed a secure system with all the necessary mechanisms, in fact we felt that we were developing a rather boring system. It is true that we came with concepts from the world of crypto, which is a 'wild west' where you can do anything and not ask permission from anyone, but it was supposed to be a phased launch. In the first phase only institutions were allowed to participate, in a limited amount, and with only one currency - the dollar. This created a conflict, because on the one hand we wanted revolutionary modernization, and on the other hand we required small steps. The process lasted almost four years, and for technology engineers it is very frustrating. It progressed very slowly."
The system may have been boring, but the huge potential bothered many.
"I don't think the concern was justified."
"Because it was a payment system, and according to the predictions of our economists, people were expected to have 20-50 dollars in their wallet. Therefore, even in the optimistic scenario, in which all two billion Facebook users would enter the system with such an amount, it is still a volume that is less than the stablecoins that exist on the market like USDC, and they didn't subvert any existing system."
"We wanted to change the world"
On July 2 everything stopped. Facebook realized that no federal regulator - and later it became clear that no state regulator either - would give them their blessing to launch the payment system. "We were told explicitly, and more than once, that we have the best stable currency system on the market," Malkhi says, "but there was a great fear of a technology giant, and especially Facebook, entering financial systems and payment systems. There was no way to break this resistance. At this stage we actually started handing over the project. We sold all its components to Silvergate, and closed Diem."
Sounds heartbreaking, after four years of working on something that stands at the top of technological development.
"Sometimes, when I introduce Diem, I explain that it was a project that aspired to get to the moon. And when you try to fly to the moon, you either reach it or explode, but on the way you light up the sky. We lit up the whole sky."
Malkhi is not exaggerating. Since the world learned about Diem, governments' interest in launching Central Bank Digital Currencies has surged. A survey among 65 central banks found that 86% are working to develop a digital currency, including the Bank of Israel, which conducted an experiment with a digital shekel. More than that, the project itself gave rise to a renewed discussion of the nature of money and raised the question of whether it is a creation of the state, which alone determines what its citizens will use, or whether it is a creation of the private sector.
Was there a point where you thought of doing it like everyone else in the crypto market, just launch?
"For what? We didn't want to launch a currency, we wanted to change the world."
Now that you are in the turbulent crypto market, which is also saturated with scams and nonsense, what do you think about what is happening in it?
"We are in an interim period. Innovative technology at this level is not perfect overnight. The key will be the bridges between it and the real world. The systems we are building will interact with both regulation and the standard financial world and will create transparency and competition. But for this to happen, we have to try and develop things that might collapse, we must optimize and remove friction from technology.
"Even in the early days of the Internet, it took time for regulators to know how to regulate, while anyone could use the technology. A similar thing will happen to crypto and efficient and safe systems will be developed, with supervision and a technological base that is much better and more transparent than what exists today."
How about private money like bitcoin, not backed by the state?
"Fiat hasn't been around for many years, it took us a long time to get there. We've tried a lot of things before, gold for example, and we know it creates an imbalance between places that have it and those that don't, and that it doesn't respond quickly enough or in sync with changes in the size of the economy - and then a crisis has arisen. Right now the Fiat system is the best and I would not replace it with something else, even if it is mathematical."
Looking back, was your project a mistake?
"I still think there is a need for a platform of the type we developed. And who will develop it, central banks? Commercial banks? Governments? No. The private market needs to introduce efficiency and innovation. Only it is capable of that."
And will it be blockchain based?
"This is currently the best solution we have to ensure that no one has control."