Machine learning — puzzling ‘big data’ nonsense
from Lars Syll
If we wanted highly probable claims, scientists would stick to low-level observables and not seek generalizations, much less theories with high explanatory content. In this day of fascination with Big data’s ability to predict what book I’ll buy next, a healthy Popperian reminder is due: humans also want to understand and to explain. We want bold ‘improbable’ theories. I’m a little puzzled when I hear leading machine learners praise Popper, a realist, while proclaiming themselves fervid instrumentalists. That is, they hold the view that theories, rather than aiming at truth, are just instruments for organizing and predicting observable facts. It follows from the success of machine learning, Vladimir Cherkassy avers, that “realism is not possible.” This is very quick philosophy!
Quick indeed!
The central problem with the present ‘machine learning’ and ‘big data’ hype is that so many — falsely — think that they can get away with analysing real-world phenomena without any (commitment to) theory. But — data never speaks for itself. Without a prior statistical set-up, there actually are no data at all to process. And — using a machine learning algorithm will only produce what you are looking for.
Machine learning algorithms always express a view of what constitutes a pattern or regularity. They are never theory-neutral.
Clever data-mining tricks are not enough to answer important scientific questions. Theory matters.
Leave a comment
—– look inside —– $5.94 / $20.00
—– look inside —– $4.90 / $8.00
—– look inside —– $15.99
—– look inside —– $5.99 / 12.99
—– look inside —– $5.93 / $12.99
—– look inside —– $4.97 / $9.90
—— Ugarteche, Puyana and Madi ——
Gerson Lima / Maria Alejandra Madi
Edward Fullbrook and Jamie Morgan
————— Michael Hudson ————–
Maria Alejandra Madi / Jack Reardon
————- Edward Fullbrook ————-
—————— Steve Keen —————–
————— Richard Smith —————
————– Gustavo Marques————
– Victor Beker and Beniamino Moro –
————– Lars Pålsson Syll ————-
—————– Stuart Birks —————-
Edward Fullbrook and Jamie Morgan
———— Armando Ochangco ———-
Shimshon Bichler / Jonathan Nitzan
————— Mauro Gallegati ————–
————— Herman Daly —————-
————— Asad Zaman —————
—————– C. T. Kurien —————
————— Robert Locke —————-
Real-World Economics Review Blog 
If it works, it works well as Facebook, Amazon, Google and others well understand. When it fails? A recent set of studies on the rise of electricity to replace fossil fuels, drive the rise of electric vehicles and related issues shows that data driven predictions have failed but then have theory driven models leading one observer, investor, to note: Bet on the Innovators and not on the forecasters. This has pointed out a major epistemological problem for those in the area of futures studies or variants such as Anticipation theory.
Does this observation also hold for both Neoclassical and Heterodox economics?
T – Great comment & issues, indeed! Yet, I wonder if the Q&As will become moot soon. It seems to me that AI is not helping any of us think about moving half of our urban infrastructure (including about 120 nuke-power plants, fuel & waste) & residents to high ground, about 129 meters higher.
Maybe I’m mistaken. Yet — despite CO2 level being over 400 ppm & increasing & methane release accelerating — I have yet to see news of anyone using ML ‘AI’ to plan for an ice-free future within the next 70 years or so.
So, my guess is that the sociologically derived “false reality syndrome” is the new global normal, explaining the absurdities & ongoing atrocities ignored by economics & most economists.
Data, whatever big or small, wherever in brains or in computers, have never been processed by themselves, but by Instructions which impose something new into data, likely that food is subject to gnawing by mouth instead of by itself. This is what all recent criticisms on quantitative analyses failed to recognize. “Theory” in the post is nothing but the result of former qualitative analyses, so what confused in the post at least lie at two aspects: now vs. before, structural vs. quantitative; Or, in another word, Roundabout Method of Production missing, pluralism missing. A new paper “The Birth of a Unified Economics” available here: DOI: 10.13140/RG.2.2.22567.29607 https://goingdigital2019.weaconferences.net/papers/how-could-the-cognitive-revolution-happen-to-economics-an-introduction-to-the-algorithm-framework-theory/ Thanks!
BLA – Thanks for the link & initiative! I find your comment’s key critiques inscrutable, but I eagerly look forward to reading about & discussing the essentials of a unified economics, especially based on a valid metatheory of real metalogical principles of unitive meta-economics or sociocultural holonomics.
BTW, do you agree with Veblen’s & Schumachers’s views of the necessities?
Sorry for delay of reply to you. I do not think Veblen’s & Schumachers’s views of the necessities are theoretically important, because what they worried or stressed are only some detailed or temporary problems which will be solved or mitigated by innovations in the long run. When economics is defined narrowly, market is obviously not omnipotent; but, when economics is extended into a unified social science, transactions are extended to various human activities (including political actions), money is extended to various adaptive measures, Veblen’s & Schumachers’s concerns would be quite more tractable — although failures and flaws would Algorithmically concur, rather than the general equilirium neoclassically.
Lars – Thanx again! Very refreshing, potent little post.
However, the ML/AI Big Data nonsense is only puzzling if we ignore the paradigm & realities of globalist ‘winning-class’ kleptocracy.
AI sys.dev & uses for fun, profit, status, illusory security & maintenance of False Reality (for as long as possible) is no more puzzling than the history of Trump, Trumpies & Trumponomics. For example, the main uses & ‘benefits’ of AI systems are perfect for automating the insanity of E. Bernays’ recipe for maintaining kleptocratic hypnocracy — until the coastal nukes start turning into our final doomdays fiasco (due to our manmade Biblical-class, post-ice global flood).
Of course, that endgame scenario assumes ongoing lack of realism & no effective action to prevent horrific dystopian doom in a toxic radioactive habitat poisoned for a million years or so.
Lars, just prior to the quote, Mayo writes this — following Popper, I assume, “The actual procedure for learning in science is to operate with conjectures in which we then try to find weak spots and flaws. Deductive logic is needed to draw out the remote logical consequences that we actually have a shot at testing (ibid., p. 51). From the scientist down to the amoeba, says Popper, we learn by trial and error: conjecture and refutation (ibid., p. 52). The crucial difference is the extent to which we constructively learn how to reorient ourselves after clashes.
Without waiting, passively, for repetitions to impress or impose regularities upon us, we actively try to impose regularities upon the world… These may have to be discarded later, should observation show that they are wrong.”
This follows from Mayo’s interpretation of Popper’s views on science and pseudoscience. “Redefining scientific method gave Popper a new basis for demarcating genuine science from questionable science or pseudoscience. Flexible theories that are easy to confirm – theories of Marx, Freud, and Adler were his exemplars – where you open your eyes and find confirmations everywhere, are low on the scientific totem pole (ibid., p. 35). For a theory to be scientific it must be testable and falsifiable.”
All well and good. And many physical scientists agree with this approach to science, including Richard Feynman. It’s also a common approach included in many cultures in the world to separate justifiable from unjustifiable knowledge. So, Popper’s views are not quite as unique as Popper sometimes claimed.
But to borrow from Shakespeare’s Hamlet, “There’s the rub.” In the world of Sapiens, and the cultures and societies Sapiens creates, including scientific culture and scientific society, people have already performed the deductive and inductive work. Already invented the theories and explanations. Already test these theories and explanations regularly. They have created the worlds they inhabit and, in doing so have created humanity. How does any social or behavioral scientist go about “falsifying” what communities of humans have created? Obviously, no scientist can carry out such falsification. But what scientists who study behavior, action, culture, and society can do is first, study these creations of humans and write about them so humans with different behavior, culture, etc. can read about and hopefully better understand “the other.” Second, these scientists can listen to the concerns of the people they study, attempt to understand the challenges and problems these people believe they face, and work with them to find ways to alleviate these challenges and problems. If this is possible. And if all involved agree with the changes in society and culture proposed to achieve these ends.
For further reading check out these,
1. Man Makes Himself, V. Gordon Childe. 1936, 1941, 1951. Begin with pages 13-19.
2. The INVENTION of HUMANITY: EQUALITY and CULTURAL DIFFERENCE in WORLD HISTORY, Siep Stuurman. 2017. Introduction.