In terms of our ability to pass knowledge on to the next generation, I think we are living in a new dark age.
Most people would say.
“We’ve never known more than we’ve known today.”
Perhaps these people have not considered that there is more knowledge floating around today than ever before but filtering that knowledge out of the cloud has never been more difficult. If we are not careful, the next generation will forget the most important knowledge because they have spent their lives mired in a fog of noise and misinformation.
Traditionally, we filter knowledge through experts – old people with loads of experience. But today, young people are bypassing those experts and deciding what they believe is true based on what they “like”.
Consider this guy, Frolly. He has an excellent overview of geological science because he has been studying it for his whole life. When he sees what the young people in media are telling everyone about how the planet works, he is appalled.
He certainly takes his time to make his points, and what is sad is that most young people today do not have the patience to follow his arguments. Their attention spans have been permanently damaged by their entertainment devices. Whereas previous generations were conditioned to see slow, calm exposition as evidence of careful, trustworthy thinking, people of today are more likely to give credence to someone with a manic, energetic, and confident presentation style.
I look around the scientific community and see young men with access to simulation tools on supercomputers telling old men that their old ways of knowing things are obsolete – we should all just trust what the supercomputer or data mining algorithm tells us, even though it has never demonstrated predictive power. When I see this breakdown of the scientific method, I know that we are in a very dark age in which the signal of knowledge is getting buried in all of the noise made by young people on the internet.
If these young data ‘scientists’ succeed in getting everyone to abandon the classical scientific method, analytic models that anyone can use would disappear and be replaced by supercomputer-based models that overfit old data and have so many input variables that they can ‘predict’ whatever the user wants them to predict. This would take the power of basic science out of our hands and sequester it in the hands of wizards at military labs like Lawrence Livermore National Lab.
I don’t want to give up the scientific method in exchange for AI data miner fortune telling that spins whatever story the user wants to hear.
I want to say to these people, “If you construct models which have not demonstrated predictive power with statistical significance and you do not warn people that your model is unscientific and unverified, then you are not a scientist. If you encourage people to take action based on such a model, you are as irresponsible as those physicists who warned all of us about a human population explosion and catastrophic collapse that should’ve happened twenty years ago but didn’t. What if an idiot billionaire had taken that projection seriously and engineered a ‘preventative’ plague in response? What if an idiot billionaire takes unverified climate projections seriously and fills the upper atmosphere with sulfur and particulates, thereby causing a famine that kills billions?”
The scientific method is the only way we know of avoiding delusion and if you do any of the following and call it ‘science’, you are dangerously deluded.
- data dredging: searching for random correlations and inferring causation and statistical significance where there is none
- p-hacking: repeating your measurement until you get the result you want and then claiming that the result is statistically significant when it isn’t
- non-falsifiability: fine-tuning an overfitted model until it matches the measurement
Real scientists know that these are cardinal sins. Just imagine if those methods were used to search for people suspected of wrongdoing and to gather evidence to ‘prove’ their guilt. Consider this quote from a scientist at Lawrence Livermore National Lab:
Scientists are much more likely to talk about “models” than “hypotheses” or “theories.” Quoting E. P. Box: “All models are wrong, but some are useful.” Because we’re focused on utility instead of correctness, we can get a lot more done without worrying if a hypothesis is “right” or “proven” or “correct.”
Simulation has upended how we think about science. The ability to take a handful of models and scale them up to run on supercomputers has opened up worlds in chemistry, physics, biology, astronomy, and engineering that we couldn’t dream about before. Of course, the individual models were “wrong,” and wrongness can accumulate as you run a few million models a few million times a second. But in a lot of fields, that’s how science is done now (and done successfully).
He has described a situation in which a model cannot be falsified. Instead, the model lasts for as long as people believe in it. That is religion, not science.
Falsifiability is the cornerstone of the scientific method. Without it, you end up calling string theory science instead of philosophy. You also end up simulating pictures of the universe or of distant astronomical objects and calling that science. It isn’t – it is art. With this style of thinking, you could start out with the hypothesis: ‘person X has a criminal temperament’ and keep digging through ancient data until you find ‘evidence’ to support that claim – even though they’ve never committed a crime.
The only valid scientific method is to make a prediction and then measure that predicted thing happening over and over again until you are sure that it is repeatable. If a computer scientist tells you otherwise, he is not a scientist. He is drunk on the power he imagines his simulation tools have given him.
The scientist from Lawrence Livermore continued:
Prior to about 1960, the data we had was data a human collected more or less manually. Now we have instruments that can generate petabytes of data over a long weekend. We’re still figuring out how to make the best use of those firehoses.
This describes a situation in which, with enough random data, you can find whatever result you want. With infinite, random data, you could find the complete works of Shakespeare. Nowadays (bad) scientists are using ‘templates’ of the signal that they want to find and then searching for it within petabytes of noise (LIGO and black hole imaging). This is very, very bad, and yet no one seems to notice, because they have forgotten the importance of the scientific method.
There is hope for us yet. Data ‘science’ high priests haven’t burned all of the witches at the stake yet.
When the internet flattened the hierarchies we’ve used to determine what is true and what is false, groupthink began to determine the truth of matters and the scientific community was not immune from this trend. Whereas Orwell imagined groupthink enforcement by the thought police (moderators), today we know that groupthink is enforced by each individual’s fear of being seen as out of step with what is popular.
For example, at the moment, it is positively criminal to criticize what “climate scientists all agree about” and it isn’t necessary for thought police to enforce uniformity of belief because we enforce it ourselves through our virtue signaling posts on the internet. With these fear-driven, virtue signaling posts, we create so much noise that debate is quickly extinguished.
“How can you not be afraid that the world is coming to an end because of how evil we all are with our carbon producing ways!?”
“Are we really smart enough to understand and control our impact? What if carbon isn’t the thing we should be worrying about. What if, by focusing on it, we are ignoring something more important?”
“I’m too busy to think about the details, but I think you are probably stupid for not believing what everyone else believes.”
For me, becoming an adult has been a process of realizing how rare truly smart people are in the world. I believe that each individual has the capacity to be smart, but smart people are rare because we are all conditioned to be so busy that we never have any time to do any real thinking. Constant distractions and striving to survive or to reach the next level of the video game are making thought impossible.
I saw a video yesterday of a group of people on a subway train singing along with a Boys2Men song from the 1990s – The End of the Road. Everyone was smiling and enjoying the sense of community the group singing produced. The song had a really simple melody, but every single voice I heard was so tone-deaf that the result was noise. I couldn’t believe it. Something has happened to our culture and our minds that is damaging on a scale that I can scarcely fathom. It is not normal for so many people to be tone-deaf. Have they been poisoned by a physical substance or by television or smartphones? This is the mystery of our dark age and until we unravel it, I think things will get darker.
I’ve been thinking about how the noise of our dark age can be explained through the evolution of music. It is a sort of zooming in process, starting back at the dawn of the enlightenment when music was about everything. It was epic and depicted the world as a whole. Then the romantic era began and the music sang the songs of the individual soul. The joy, the despair, the currents of mood flowed in sweeping melodies. Then the world wars started and the music became focused even further with syncopated klinks of the keys depicting the zips and zaps of the currents scientists told us flowed within our minds. The meaning could not be deciphered, but it depicted an aspect of our selves. After the world wars, the music got louder. If you take the words away, you have the banging and booming sounds, the rock and the roll of abstraction which has evolved into the pulsing heartbeat of electronica and a cacophony of white noise. This is the song of our dark age, but there are hints of a new enlightenment dawning. We will eventually emerge from an intellectual bottleneck that selects new, epic songs to describe the world as a whole.
I first posted this material on quora.com