If you want to make sure that a person or a quantum computer is telling you the truth, you either have to convince them that you already know the truth and that you have information that could hurt them… or you might have the person or quantum computer record what it plans to do and then check that it did what it said it was going to do.
This was the great epiphany on which a graduate student named Urmila Mahadev spent almost a decade of her life. The reason that it took her so long to have this epiphany was that she needed to figure out how to say these things in a coded, academic language that would be very difficult for non-specialists to parse.
She had a computer that could calculate something but she had no way to know if it had calculated that something correctly because there were no other computers that could perform the calculation in a verifiable way. Her computer could perform the calculation a million times and get the same answer each time, but that isn’t the same thing as knowing that the answer calculated is correct.
Her job was to convince everyone that there was a way to be sure that the answer was correct.
She did this by convincing herself and others that if you perform a calculation with a known outcome in parallel with a calculation with an unknown outcome, a correct result from the known outcome calculation will validate the result from the unknown outcome calculation.
I’m not sure that I believe her result.
Can you really be sure that the quantum computer delivered a correct result for the unknown calculation? I don’t think you can and I don’t believe she solved the problem. I think she just formalized and encrypted the statement of the problem and solution in language that was sufficiently complex to trick people into believing that the problem has been solved. She wrote a proof by tedious complexity which, incidentally, might be good enough to solve a real problem that needs to be solved: how do we convince the world that the government has the world’s most powerful encryption tools? How do we convince the world that resistance is futile?
I congratulate her for her perserverance. The educational system is designed to break your mind and the fact that it took almost ten years for them to break her ability to think or communicate like a normal person should make her parents proud.
She has learned to speak in a quantum, academic code and lost her ability to use plain speech. In other words, she has learned how to say things in a way that will be indecipherable to most people — even, in some cases, to herself. At the end of the breaking process, many a young physicist will not even realize that he or she has become indecipherable.
Those who have not undergone this process of encryption or indoctrination have usually retained their ability to think in an uncoded, clear, rational fashion. To them it requires no lengthy proof to see that the best way to encrypt something is to bury the message in irrelevant noise and to use extended euphemisms that require special, cultural knowledge to decode.
I believe that the journalist who wrote the article about Dr. Mahadev was well versed in that form of encryption because his article read like an extended euphemism. Of course, to those who lack the ability to think about things figuratively, anyone who insists that something is a euphemism will be seen as crazy. Therein lies the power of this form of encryption. It has built-in plausible deniability, unlike a code in which the speakers simply randomly mis-spell words here and there to mess up a data miner’s ability to extract meaning.
I believe that the journalist at Quanta Magazine indirectly told readers with the proper cultural context (decryption keys) that the power of the quantum computer is not in what it can actually do — it is in the threat of what it can do. In many ways, it is like the Star Wars space program.
If you are not up to date on your history, the US military convinced Russia that it was building free electron lasers in space so that it could shoot down intercontinental ballistic missiles. This program was called Star Wars. Russia bought the bluff and this somehow changed the course of history. Of course, that was the US narrative. I’m sure that Russia’s narrative would be quite different.
Today, we have a quantum computing community that claims the ability to use fifty three qubits to perform an unverifiable quantum calculation over and over, a ridiculous number of times while getting the same answer each time. The calculation itself has no known applications, but it supposedly would’ve taken a conventional computer 10,000 years to complete. They plan to scale this up to thousands of qubits without knowing how they would use such a system.
This brings us back to the topic of Dr. Mahadev’s thesis. If you can’t verify the results of these computations, how can you be sure that the calculation was correct, even if it was done thousands of times and got the same result? If there was a mistake anywhere in the chain of logic, the answer would be wrong. She believes that if you embed a verifiable calculation within the unverifiable calculation, then the unverifiable calculation will be verified. I’m not sure about that.
I suppose that you could have a network of quantum computers that all know the right answer and check one another, but that doesn’t change the fact we wouldn’t know the right answer and we would have no idea if the computation was of anything that has any connection to objective reality. If the result has no connection to objective reality, it cannot be used, for example, for physics calculations or for simulations of economics or social dynamics.
If you’ve read my piece on quantum computing, you’ll understand my scepticism about the usefulness of the method. A system that can only compute simulations of itself isn’t terribly useful. Then again, isn’t that what a conscious being is doing all of the time? Maybe a network of quantum computers that evolves in response to a threat might serve a purpose, but that calls to mind some frightening science fiction scenarios.
Anyone in academia knows that students are trained to encrypt their knowledge so that they can communicate with a community of elites while excluding the hoi polloi with their jargon. Once a student has demonstrated adequate encryption skills, he or she might be allowed in the room with the elders of the community. But even then, they have to toe the party line. Since my experience has been limited to that of government research labs, I wonder if this behavior is unique to the community of scientists. I certainly see this behavior within the community of data scientists through their use of Bayesian statistics instead of more transparent frequentist analysis, but is this opacity in the interest of society?
Protecting non-democratically elected enclaves of power isn’t always in the public interest. It prevents the system from self-correcting and might necessitate violent dismantling of the enclaves when they become too powerful. The worst events in human history have stemmed from fearful people in powerful enclaves who have a vision of the world which is warped beyond belief. When, through their fear-warped vision, they see a flawed system and a world heading for destruction, they may be inclined to set the world on fire so that they may start with a clean slate and realize their twisted dream of re-shaping the world in some ideal image. When this happens, mass murder or suffering has typically followed. History has shown us that the ends do not justify the means and these fearful people’s choices operate according to zero-sum rules that do not benefit humanity as a whole.
I didn’t finish watching these guys chat, but their discussion of how the world was designed to hold half of a billion people sent chills up my spine. If someone with power truly believed that and engineered a plague to bring those conditions about, let’s just say that every atrocity in human history would pale in comparison.
When I see people with means thinking in those ways, I wish for belief in a higher-power to make a resurgence.
I wish for people like Dr. Mahadev to be broken free from their academic trances so that they can avoid becoming a cog in a machine of destruction and oppression and so that they can see what their leaders are really thinking about the future and help them achieve a less fearful perspective.
Non-deities like me and Dr. Mahadev are trained to absorb complexes of ideas and we are told that they will give us superpowers which will allow us to make a contribution to civilization. What is sad is that in most cases, this not true. This absorption was a pointless exercise that makes many people miserable. I was most miserable when I was studying physics. I would go to bed every night wishing for an earthquake to bring the world crashing down around me so that I could walk out into the world as free person. It was only decades later, after disengaging from this system that I saw how it was designed by fearful people to absorb the attention of young people that they feared. They were afraid that young people would do something disruptive and they told themselves that the encryption fostered by the system was the medicine that the system needed to survive, but I feel like they were so blinded by and driven by fear that they could not see the ways in which they were creating a poisonous environment that aggravated the condition they sought to cure. I’m not sure about any of this, but this was my gut reaction today.
If you would rather listen to this post than read it, please try the video below.