Monday, April 16, 2018

IBM has a female blockchain team

I have criticized IBM for over-hyping the potential for quantum computing and related technologies, but what it is doing with the Bitcoin blockchain is even worse. reportsL
In the male-dominated world of cryptocurrency, IBM is going against the grain. The company's 1,500 member blockchain team is led by Bridget van Kralingen, senior vice president of global industries, platforms and blockchain. Meanwhile, the actual blockchain development was led by IBM Fellow Donna Dillenberger. ...

However, the company's predominantly female leadership lies in stark contrast to other blockchain start-ups, which are overwhelmingly run by men. It also makes the company an anomaly within a fintech industry that remains heavily male-dominated. However, Kralingen, who joined the company in 2004, notes that promoting women to leadership positions isn't a new trend at IBM. The company's CEO, Ginni Rometty, is the most visible example.
The demise of IBM is sad. It once owned the whole computer market, and had no serious competition in mainframe computers. If the management had any sense at all, it would now own the cloud computing market. But it is not even a serious player.

IBM got rid of its best operations, and brought in female management. So what do they do? Jump on a stupid fad, and try to use it for social justice!
Dillenberger notes that blockchain has many uses outside of cryptocurrency and she highlights food safety as an example. Using a cellphone, a farmer can scan the exact moment a food is planted, harvested, packaged and distributed, onto the blockchain platform. This comes in handy when there's a food recall because a company can quickly pinpoint where things went wrong. This leads to greater transparency and trust among businesses and consumers, explains Dillenberger.

In fact, it's this social justice aspect that makes blockchain development so appealing to her. Dillenberger points to the 2008 Chinese milk scandal, in which six infants died because plastic was added to baby formulas and milk. She says that tracking food production through blockchain would help companies and the public avoid these types of scenarios.

"I'm an IBM Fellow," she adds. "I'm expected to push the boundaries of the frontiers of science and math and technology."

Lack of women in blockchain
However, most women are not given this opportunity. An analysis of the top 50 blockchain companies found that just 16 percent are founded by and/or led by women.

The reasons behind the gender disparity are varied, but women overwhelmingly point to a "blockchain bro" culture that caters exclusively to men. In January, the North American Bitcoin Conference featured 84 male speakers and three women, while the post-conference networking party was held at a Miami strip club.
This is crazy. The blockchain does not run on a cellphone. China could have apps for farmer to log data from a cellphone, if it wanted to. Logging data into a database is old technology that does not need a blockchain.

The blockchain lets competing "miners" do million-dollar computations in a race to time-stamp the data, so that an outside party can choose to trust the time-stamp with the most computation. (Or more precisely, the first one to meet the computational requirements.) That's all. It is not going to stop a cheating milk producer from contaminating milk. Logging the supply chain might help trace a problem, but the blockchain itself does not add anything useful.

IBM has 1500 people working on this scam!

The major blockchain startups got the big funding about 5 years ago. There has been plenty of time to see a product by now. And yet there is still nothing useful.

Friday, April 13, 2018

Quantum random number generator is bogus

A new Nature article (paywalled) claims to have invented a random number generator:
Researchers have come up with a way to generate truly random numbers using quantum mechanics. The method uses photons to generate a string of random ones and zeros, and leans on the laws of physics to prove that these strings are truly random, rather than merely posing as random. The researchers say their work could improve digital security and cryptography.

The challenge for existing random number generators is not only creating truly random numbers, but proving that those numbers are random. "It's hard to guarantee that a given classical source is really unpredictable," says Peter Bierhorst, a mathematician at the National Institute of Standards and Technology (NIST), where this research took place. "Our quantum source and protocol is like a fail-safe. We're sure that no one can predict our numbers."
He is "sure", because it is "like a fail-safe"?!

No, they are making a fundamental conceptual error here is talking about "truly random numbers", and thinking that quantum mechanics is some sort of magic ingredient.

There are lots of commercially available random number generators, whose output no one can predict. There is probably one on whatever computer you are using to read this message. See RdRand for details.
"Something like a coin flip may seem random, but its outcome could be predicted if one could see the exact path of the coin as it tumbles," adds Bierhorst, "Quantum randomness, on the other hand, is real randomness. We're very sure we're seeing quantum randomness because only a quantum system could produce these statistical correlations between our measurement choices and outcomes."

When it comes to the source of the randomness, we're back to good old quantum superpositions, where a quantum particle can be a one, zero, or both at once. The measurement of these superpositions has fundamentally unpredictable results. The method doesn't so much use the superpositions themselves to generate the data, but the correlations between these superpositions when photons are looked at in pairs.
No, this is nonsense. Quantum randomness, if there is such a thing, probably affects coin tosses also, and they are not predictable.

Quantum randomness is often emphasized in the Copenhagen interpretation, but those who follow other interpretations, such as pilot wave, many worlds, and super-determinism, deny that there is any such thing. If we could prove quantum randomness, then we could eradicate these other silly interpretations. Unfortunately, we cannot.
However, the team goes a step further to improve the quality of its data. By analyzing the data produced, the researchers can home in on shorter strings where the occurrence of ones and zeros is nearer to fifty-fifty. The team has written a computer program to select the strings, which ironically uses a conventional random number generator to provide seed data, effectively telling the program what to look for.

The researchers call this proximity to fifty-fifty perfection "uniformity." From the more than 100 million bits generated, the researchers found 1,024 certified to be uniform to a trillionth of a percent. "A perfect coin toss would be uniform, and we made 1,024 bits almost perfectly uniform, each extremely close to equally likely to be 0 or 1," Bierhorst explains.
Okay, this is so stupid that it appears that we are being trolled.

They make this perfect quantum truly random number generator, but then they scanned the output to pick out the more random-looking numbers?!!

It is well-known that people are lousy at detecting random sequences, because they expect too much uniformity. See How to tell whether a sequence of heads and tails is random for a simple explanation. But these authors have rigged their system to get greater uniformity.

A advantage of the mathematical pseudo-random number generators is that they automatically get the uniformity correct. It is inherent in the generating formulas. A particular output may look non-uniform, but that is supposed to possible in a random number generator.

This paper was published in one of the world's top two scientific journals, as a great advance in quantum randomness or whatever the subject is. It is not. As a random number generator, it is not as good as what is currently in use in billions of computers. As a paper in quantum mechanics, it is hopelessly confused.

Sunday, April 8, 2018

Worldview underlying blockchain is wrong

The Bitcoin/blockchain hype continues, impervious to its history of failure.

Kai Stinchcombe rants:
Blockchain is not only crappy technology but a bad vision for the future. Its failure to achieve adoption to date is because systems built on trust, norms, and institutions inherently function better than the type of no-need-for-trusted-parties systems blockchain envisions. That’s permanent: no matter how much blockchain improves it is still headed in the wrong direction. ...

There is no single person in existence who had a problem they wanted to solve, discovered that an available blockchain solution was the best way to solve it, and therefore became a blockchain enthusiast. ...

So in summary, here’s what blockchain-the-technology is: “Let’s create a very long sequence of small files — each one containing a hash of the previous file, some new data, and the answer to a difficult math problem — and divide up some money every hour among anyone willing to certify and store those files for us on their computers.”

Now, here’s what blockchain-the-metaphor is: “What if everyone keeps their records in a tamper-proof repository not owned by anyone?” ...

The entire worldview underlying blockchain is wrong

You actually see it over and over again. Blockchain systems are supposed to be more trustworthy, but in fact they are the least trustworthy systems in the world. Today, in less than a decade, three successive top bitcoin exchanges have been hacked, another is accused of insider trading, the demonstration-project DAO smart contract got drained, crypto price swings are ten times those of the world’s most mismanaged currencies, and bitcoin, the “killer app” of crypto transparency, is almost certainly artificially propped up by fake transactions involving billions of literally imaginary dollars. ...

Even the most die-hard crypto enthusiasts prefer in practice to rely on trust rather than their own crypto-medieval systems. 93% of bitcoins are mined by managed consortiums, yet none of the consortiums use smart contracts to manage payouts. Instead, they promise things like a “long history of stable and accurate payouts.” Sounds like a trustworthy middleman!
He is right. It is nearly impossible to find a problem for which a blockchain is an appropriate technological solution.

The country of Estonia has a system it used to call hash-linked time-stamping, and now calls it the Estonian blockchain. It is useful, but it is not a blockchain, and it even predates the invention of the Bitcoin blockchain.

The blockchain only makes sense if there is a public ledger, and there are competing players willing to spend unlimited resources to keep up with each other to maintain the integrity of the ledger. Furthermore, you have to have a user base that does not trust any of the players in particular, but does trust their ability and willingness to competitively carry out joint authentication operations, without any player outworking the others.

The Bitcoin blockchain does have some utility in illicit money transfers to the other side of the world. Are there any other examples? You might read of applications like tracking the WalMart supply chain, but such applications don't make any sense.

Friday, April 6, 2018

TED Talk promotes quantum crypto

A recent TED Talk was mainly a snake oil pitch for his company's useless products:
How quantum physics can make encryption stronger | Vikram Sharma
Tuesday, March 27, 2018, 7:58 AM

As quantum computing matures, it's going to bring unimaginable increases in computational power along with it -- and the systems we use to protect our data (and our democratic processes) will become even more vulnerable. But there's still time to plan against the impending data apocalypse, says encryption expert Vikram Sharma. Learn more about how he's fighting quantum with quantum: designing security devices and programs that use the power of quantum physics to defend against the most sophisticated attacks.
He starts with scare stories about data breaches, including saying that the cyberthreat is now affecting our democrat processes because some Democrat National Committee emails were stolen.

Except that we don't actually know that a cyberattack had anything to do with those emails. It was widely reported that 17 intelligence agencies looked at this, but in fact none did, as the DNC refused to let the FBI look at the servers, presumably because they contained incriminating data.

Many also believe that DNC insider Seth Rich leaked those emails to WikiLeaks.

It also appears that these leaks improved our democratic processes because they exposed primary favortism and fundraising collusion within the DNC.

And his companies products would not be any help. He brags about a product that generates "true random numbers" by hardware, not software. And he raves about the potential of quantum key distribution.

His products and plans are nearly worthless. Random numbers are not hard. The following method has been around for 25 years or so. Flip a coin 160 times. Apply SHA-1 repeatedly to this bit string followed by a counter, to generate all the random numbers you want.

Quantum key distribution doesn't really solve any problems, because you need to replace all your routers with quantum computers, and because you cannot authenticate anything, and because it is nearly impossible to make equipment that matches the theoretial models.

He also repeats this nonsense that physical assurances of security are somehow better than mathematical assurances. He and others in this field like to say that they are relying on the laws of physics to be truly uncrackable, instead of math-based cryptography that has shown to be fallible again and again.

I can't think of a single example of a business or orgranization that suffered some loss because of a break in math-based cryptography, when the system was following generally accepted best practices. That goes for DES, RSA, SHA, DSA, ECDSA, etc. Systems have been broken because of bugs and implementation flaws and even hardware failures, but not from breaking the math.

On the other hand, the quantum key distribution devices have all been broken.

QKD theory will make assumptions like the device emitting a single photon with a particular frequency and polarization, and a detector will measure that photon's polarization. This sort of precision is physically impossible. You can emit light that is probably 0, 1, or 2 photons in approximately the right color and orientation. But you need special info that might leak info in a hardware attack.

Monday, April 2, 2018

The embryonic stem cell research failure

I criticize quantum cryptography and quantum computing for being over-hyped vaporware. But there was another technology that was overhyped about 100x as much about 10-15 years ago - embryonic stem cells.

It was so hyped that a California ballot initiative funded it with $3B. Professors everywhere said that it was a reason to vote for Barack Obama.

Nature mag reports:
... in 1998 when researchers first worked out how to derive human embryonic stem cells. ...

Starting with an attempt to repair spinal-cord injuries in 2010, there have been more than a dozen clinical trials of cells created from ES cells — to treat Parkinson’s disease and diabetes, among other conditions. Early results suggest that some approaches are working: a long-awaited report this week shows improved vision in two people with age-related macular degeneration, a disease that destroys the sharpness of vision.

“In some ways, it’s not a surprise, because 20 years ago we expected it,” says Egli, “but I’m still surprised that this promise is becoming a reality.”
Everyone in this field was publicly bragging about much more dramatic progress. They were saying that after about 2 years of research, paralyzed people would be walking again.
In 2001, US President George W. Bush restricted government funding to research on just a few existing ES-cell lines.
More precisely, he expanded funding to research on about 50 lines, after Pres. Bill Clinton had banned funding entirely. Pres. Obama later expanded it to about 60 lines.

Nature calls this progress a "revolution", but it is pitiful compared to the tens of billions in funding, and the wild promises of miracle cures. Here we are, 20 years later, and the only clinical benefit they have to show is that two people have some improved vision. It doesn't say how many patients got worse vision from the experiment. Nothing has come out of that California $3B, as far as I can see.

As I write this, I am listening to a recent interview of Sam Harris. In a rant against religion, he complains:
they want to throw gays off of rooftops, force women to live in bags, or prevent gay marriage in our context, or prevent embryonic stem cell research

Stem cells seem to be a leftist-atheist bugaboo. Obviously these ideas have some sort of symbolic significance to him that rivals that of the religious folks he attacks.

So why is there all the over-the-top hype? You could just as well ask why the leftist-atheists were so preoccupied with gay marriage. I think they see it as invading God's turf.

After the article called it a "revolution", I expected it to be called a "paradigm shift" also. Those are the terms for overhyping advances that are not really advances. For example, string theorists are always talking about revolutions, even tho they have never made any real advances.