https://www.getsmarter.com/courses/us/mit-blockchain-technologies-online-short-course Dear Colleagues, I noticed that the site GetSmarter.com is announcing a course on "blockchain technology" that would be offered by your prestigious School. Please reconsider and exclude that topic from your curriculum. Blockchain is a computer science version of the old snake oil elixirs. It is not a new, revolutionary, or promising technology, not by a long shot. It is a buzzword without any technical merit. Seeing a course on that topic offered by your School is like seeing a cource of crystal healing offered by a top Medical School. Blockchain technology is touted as a way to produce a "decentralized" system: one that can be trusted without the need to trust any specific participant. But that claim is false. Proponents of the "technology" distinguish two types of blockchains, "permissioned" and "permissionless". The former are in fact centralized, while the latter are impractical. The "permissionless" blockchains incorporate the key idea of bitcoin, namely that anyone is allowed to update the data structure, anonymously and with no need for prior registration or authorization. A system or rewards and a highly expensive "proof-of-work" system are claimed to motivate the updaters ("miners") to work "honestly" and to reject malicious updates, as well as prevent the so-called "Sybil" attacks in which malicious updaters are cloned by the thousands. However, a permissionless blockchain needs a cryptocurrency in order to reward the "miners" for their work. That in turn requires a large contingent of dumb "investors" who, by buying the coins that the miners create, will pay the cost of their proof-of-work effort -- even though there is no source of money that could return to those "investors". And the cost of running a permissionless blockchain *must* be huge, because it is what is supposed to make the chain secure against tampering. Basically, if tampering with a piece of information in some block could give 100 million dollars of profit to someone, that information cannot begin to be considered secure until the miners have expended at least 100 million dollars worth of electricity to create blocks after that one. (At present, the bitcoin (BTC) miners take from investors about 50 million USD each day, and consume at least half of that in electricity. Their total consumption is estimated to be at least 7 gigawatts of electric power. That is the output of seven standard nuclear reactors; more than the electricity consumption of Austria, and almost three times that of Ireland. Still, it would take a week before a bitcoin payment worth 100 million dollars could begin to be considered secure against fraudulent reversal.) In the "permissioned" blockchains, on the other hand, there is no proof-of-work requirement. Instead, the miners are selected by some central entity or consortium, that vets for trustworthiness, watches their behavior, and excludes any which are not working "honestly". But then the only alleged advantage of blockhains, namely decentralzation, is lost. That central entity can command the miners to rewind the blockchain and reverse transactions, and exclude any that fail to obey. Blockchain technology proponents incorrectly claim that the concept was invented by the creator of Bitcoin, the first "cryptocurrency". But a blockchain is basically a log of events, divided into blocks, each cryptographycally stamped in such as way tha any change to the information in a block requires rebuilding all subsequent blocks. This type of data structure is in fact at least 30 years old. Due to its sstructure, the only approriate role for a blockchain in an information processing system or network would be that of a log of events. But well-designed systems already should have such logs, for secondary purposes such as auditing, archival, statistics, analysis and recovery from software bugs, and synchronization of gegraphically separated servers; and those logs are often organized and cryptographically stamped in a blockchain fashion. As a data structure, a serial log of events is inherently too inefficient to be used as the primary operational database of any large information processing system. The servers of a bank, for instance, would work primarily with *mutable* database that has the *current* balance of each account. Indeed, information processing professionals would not even call the blockchain structure a "database". Bitcoin's creator did not choose that "database structure" because it excelled in some quality, like efficiency, response time, security, etc. He choose it because it was the only structure that could be updated by a collection of anonymous uncoordinated and unvetted agents without the need for a central authority. The price for that feature was huge: it made the performance of his system abysmal in all those respects. More generally, the technology to implement distributed but centrally managed databases that are efficient, secure, and as reliable as one might want has been developed and used for more than 40 years. Unfortunately, most "blockchain experts" seem to be wholly ignorant of that technology, or even of the nature and actual processing problems of the applications that they set out to "improve". Blockchain technology would never have been a thing without the "success" of cryptocurrencies (``cryptos'') as speculative investment. There is a symbiotic relationship between promotion of the latter to investors and promotion of the former to IT managers of corporatons, especially in the financial industry. Crypto promoters have been pushing hard for the twin misconceptions that investing in cryptos is smart because the technology is revolutionary, and that the technology is powerful because it made cryptos possible. In reality, economists and savvy investors have long realized that crypto investing is a ponzi scheme, by now the largest in history; and practically all computer scientists and competent financial software developers can see that the technology is worthless. And, even if the separate claims were true, there would be no connection between them. Because of these obvious problems, and of the close connections of cryptocurrencies with crime, it is highly inappropriate for universities like MIT to lend their prestige to these two twin scams. Proponents of the crypto ponzi often cite those courses as indirect "proof" that cryptocurrencies are a legitimate investment. I urge the Board to look at this problem, and handle courses on blockchain or cryptocurrency as it would handle courses on astrology or MLM management. Thank you for your attention, ??? is being used by incompetent or unscrupulous entrepreneurs to obtain investment money from companies who do not Its flaws were not noticed at first, but they became obvious more than eight years ago. Simply put, they are utterly inadequate as either alternative forms of money or as systems for legal payments. Cryptocurrencies are still alive today for two reasons only: (1) They are a convenient money laundering tool for all sort of illegal activities, including traffic of drug, sex slaves, child porn (and child sex slaves), ransomware, tax evasion, bypassing sanctions, and all sorts of scams. (2) They are one of the largest financial scams in history, that already transferred some 15 billion USD from naive "investors" to the organizers (the so-called "miners"); money that will never return to the former, in any guise. In (2), like in any pyramid investment scheme, the profit of investors and organizers depends entirely on convincing more people to invest in the scheme. Thus bitcoin pushers are always trying to insert their misleading marketing into all sort of contexts, especially in those that could give the concept an aura of respectable technology and/or investment. That, unfortunately, is what seems to have happened in your organization. Sincerely, --jorge -- Jorge Stolfi Full Professor/Professor Titular Instituto de Computação/Institute of Computing UNICAMP