by Summary:
From the Climate and COVID and Climate Correlations, the New Economics (2021) of Professor Steve Keen is the topic of this weekly series. Minsky software walkthroughs are common, so come by, we want to hear what you have to say. We are always looking for new and novel ways to look at a complex subject or to the issues that matter to us most. This video is about all of this and more.
Topics:
Steve Keen considers the following as important:
This could be interesting, too:
From the Climate and COVID and Climate Correlations, the New Economics (2021) of Professor Steve Keen is the topic of this weekly series. Minsky software walkthroughs are common, so come by, we want to hear what you have to say. We are always looking for new and novel ways to look at a complex subject or to the issues that matter to us most. This video is about all of this and more.
Topics:
Steve Keen considers the following as important:
This could be interesting, too:
John Quiggin writes The war to end war, still going on
New Economics Foundation writes Reclaiming our regions
New Economics Foundation writes New Economics Podcast: Why is the benefits system failing disabled people
Michael Hudson writes Jill Stein: Splitting the Pro-Imperial Vote
| From the Climate and COVID and Climate Correlations, the New Economics (2021) of Professor Steve Keen is the topic of this weekly series. Minsky software walkthroughs are common, so come by, we want to hear what you have to say. We are always looking for new and novel ways to look at a complex subject or to the issues that matter to us most. This video is about all of this and more. |
Cheers for the stream professor!
Loved the conversational nugget of classical sciences as two body solutions in contrast to fully modern scientific approaches grappling with n-body.
@7:10 in complex systems some scaling laws are linear: e.g., water consumption vs city pop. size (if I recall correctly from Geoffrey West's work at SFI). So there are some sub-systems that linear optimization methods etc will work on, but they are the exception rather than the rule.
Love the comments on modern agriculture turning fossils into food (around 33:20). Without fertilisers from the Haber-Bosch process another energy input would be needed to support high-yield agriculture, the obvious one is human labour. A return to labour intensive agriculture could achieve sufficiently high yields but would mean a large fraction of the population working the land. Allotments often achieve higher yields than industrial extensive monoculture farming but with orders of magnitude more labour input per hectare.
@59:00 George dude, attractors do not have an "ontology." They exist in pure mathematics – and they only exist when you have real numbers (otherwise you will get ergodic recurrences). In ontology we are dealing with physics, and real numbers are in physics domains mere fictions, and given a long enough time Poincaré recurrences will occur, but only if a system is completely isolated and undisturbed. So in the real world any puff of wind or someone sneezing or a cosmic ray will shift the dynamics onto a new path. That's when you see the attractor, since the new path will look the same as the old, qualitatively, if the dominant control variables put the system on an attractor. The attractor itself is always a mathematical idealization, a Platonic essence, if you like, so it has no ontology (unless you are a strong mathematical platonist).
As a physicist I would just say if you are seeking an ontology for chaotic attractors then you will find them in the fundamental forces of constraint acting on the system: mostly gravity or EM, along with the charges they act upon which you, or nature, put into the system by preparation. There is nothing else to it. Steve Keen's Minksy module that ran the Lorenz model has that exact strange attractor, it is illustrated in pixels on his monitor screen. But there are no forces and charges inside his computer producing it directly, it is all engineered in C++/Java code at a software level, showing you the attractor is a platonic essence, not an ontological thing.
@45:30 bit of an exaggeration here. Bernoulli's Fallacy by Aubrey Clayton is about why frequentist methods are generally inferior to Bayesian methods. But for pure mathematicians they are equivalent. It is only when applied to the real world, where — as Steve points out — one cannot conduct infinitely many perfectly repeated trials, that the Bayesian analysis can be superior. But, a Bayesian has to then input priors, and these are not well-determined, they involve guesswork or at best subjective assessments. Point is, in some applications of statistical methods one needs this type of method to avoid gross errors in the tail risks. Nassim Taleb is also championing thought along such lines.
But mathematically there is nothing wrong with standard statistics. An analysis assuming infinitely many trials is perfectly in accord with Bayesian analysis when uncertainties are appropriately included, and that turns out (due to error propagation) to be mathematically equivalent to using Bayesian priors. So there really is no "crisis" in mathematical statistics. The crisis is in the social sciences where scientists blindly use statistical software like S-Plus and SPSS as a black box, and do not account for uncertainties, or use incorrect Bayesian priors. (HINT for SteveK: maybe do not use those R^2 values in your Minsky presentations. You want to measure correlations between private debt and GDP etc using mutual information or metric variance information estimators.)
I repeat: the problem is misapplication to the real world. The whole "p-value industry" and "R^2 industry" in social science for example. I highly recommend anyone in the social sciences or complexity studies to seriously consider Taleb's take-downs of p-values and correlation coefficients. If you want to assess how closely two factors are related you want to use mutual information estimators, not R^2 values. I've collected a brief list of superior techniques in the reply below for those working in these fields of research. Also, once you get into it, it is very satisfying to see how your improved analysis matches against the dumb standard use of canned stats packages. Even a non-statistician can appreciate the difference.
Thanks Bjiou. As I think you know, stats is my weak point in mathematics. Can you point me to papers on mutual information estimators, including how to code them?
In general, I use correlation coefficients in my graphs simply because, according to conventional economic theory, the things I'm correlating (credit change & asset price change, unemployment and credit, etc.) should have no significant correlation. Clearly they do. Since that aspect of standard statistics is well understood, I'm willing to continue using CCs despite the deficiencies, but if I wanted to make a causal argument, then of course they're insufficient.
book The Knowledge We Have Lost in Information: The History of Information in Modern Economics
Jun 1, 2017
by Philip Mirowski, Edward Nik-Khah
book
Untote leben länger: Warum der Neoliberalismus nach der Krise noch stärker ist (German Edition)
Sep 1, 2015
by Philip Mirowski, Felix Kurz
book
The Birth of the Business Cycle (RLE: Business Cycles) (Routledge Library Editions: Business Cycles)
Mar 27, 2015
by Philip E. Mirowski
book
Never Let a Serious Crisis Go to Waste: How Neoliberalism Survived the Financial Meltdown
Jul 9, 2013
by Philip Mirowski
book
The Reconstruction of Economic Theory (Recent Economic Thought, 8)
Nov 30, 1986
by Philip Mirowski
book
Machine Dreams: Economics Becomes a Cyborg Science
Dec 3, 2001
by Philip Mirowski
book
More Heat than Light: Economics as Social Physics, Physics as Nature's Economics
(Historical Perspectives on Modern Economics)
Nov 29, 1991
by Philip Mirowski
book
Science-Mart: Privatizing American Science
Apr 1, 2011
by Philip Mirowski
book
The Road from Mont Pèlerin
Oct 19, 2009
by Philip Mirowski, Dieter Plehwe
book
The Effortless Economy of Science? (Science and cultural theory) by Philip Mirowski
Jul 21, 2004
book
Science Bought and Sold: Essays in the Economics of Science
Jan 3, 2002
by Philip Mirowski, Esther-Mirjam Sent
book
Against Mechanism: Protecting Economics from Science (SUNY Series in Philosophy and Biology) by Philip Mirowski
Jan 27, 1992
42:41 – Deliverability of the message
31:48 –
33:05 – [. The method for producing hydro from hydrocarbons is known as steam reforming. Hydro is then combined with nitro to produce ammonia via the Haber-Bosch process]