Curiosity is bliss

Kirill Osenkov's pet project, Live Geometry, is a cool Silverlight application which lets you build geometric constructions and interact with them.
In his words, it's an "interactive designer for ruler-and-compass constructions".

Watch the 5 minute screencast and try the application on livegeometry.com.

Also, check out the class diagram illustrating the design (dependency tracking and change propagation, similar to a spreadsheet software and reminiscent of reactive programming).

PS: I wish more screencasts were that condensed. Video on the web is too slow ;-)

Every week Wired Science (on Hulu) covers a wide variety of scientific and technologic topics. The episodes I watched so far are pretty impressive and educative, especially when it comes to recent advances in medicine.


Episode 9 (seeing through the tongue)
Based on the work of Dr Paul Bach-y-Rita on brain plasticity (the ability to adapt and re-configure itself), researchers are now able to give blind patients some limited ability to "see".
Although the signal is input through a device on their tongue (like a matrix of pixel) it actually goes across senses and triggers the visual cortex. As he observed, "You don't see with the eyes. You see with the brain."

Similarly, for a patient that lost her sense of balance (after the little hair cells in her ear were damaged) a similar device allows to substitute a natural sense with an artificial one. The surprising part is the residual effect: the patient keeps a feeling of stability when not using the device.


Episode 8 (electrodes in the brain)
One of the segments reveals how some Parkinson's disease symptoms can be improved. A former patient and one about to undergo the surgery show how their hands were shaking uncontrollably, due to an area of the brain firing random signals.
Although doctors don't exactly know why, they found that implanting pulsating electrodes (like a pacemaker) in that area of the brain limits the symptoms and improves motor control.

Aside from the before-after results, the awake surgery is fascinating: to find the right area of the brain, a probe is used, which listens to neurons and you can literally hear the area of the brain corresponding to the patients hand.


Episode 1 (robot-assisted surgery)
This episode demonstrates a robot-assisted heart surgery. The surgery is much less intrusive and yet gives the surgeon rich manipulation capabilities, as the robot tools have similar mobility and dexterity to human wrists. This allows for fine suturing in enclosed spaces. But the robot itself, aside from costing a fortune, is also still quite bulky.


Episode 2 (design and ingenuity for the poorest)
An interview with Cameron Sinclair about architecture for humanity (network for designers to share ideas and solutions) and the Design like you give a damn book, with many innovations for the poor.
He gives a few examples, such as houses made from long sandbags coiled into a dome shape. These shelters are very cheap (cheaper than a refugee tent), more durable and require very little goods to be transported.
New UN tent, which is light, uses a smaller surface areas, pops out and has a divider for privacy.

The Grip Clip, a clever piece of plastic which is very versatile for assembling tents from local materials. It fastens the membrane to the support structure. That said, in an episode of Man vs. Wild, Bear Grills was showing an equivalent solution using a round stone and a string.


Along the same lines, there are great examples of entrepreneurship which can improve the conditions of life for poor farmers over the world: windmill electricity and water pump (amazing story of a 14 year old kid, pictured here), affordable drip irrigation systems in India (targeted for a 1$/day families with small fields, yielding 2x or 3x income increase to 300.000 farmers and resilience to dry seasons), cheap and local bed nets factories, cheap private healthcare (maternal hospitals, ambulances).
Those last projects are done as a mix of charity and market solutions (philanthrocapitalism) to lift people out of poverty.

Mainstream economist and textbook author Greg Mankiw notes the current unemployment metrics and reflects on the accountability and effectiveness of the stimulus (see udpated graph).
This again raises the question of methodology in economics and the necessity for the public, politicians and economists to distinguish economic science from economic forecasting.

In light of the 9.7% unemployment rate in August, Mankiw looks back at the administration's argument for stimulus in January 2009 and the expectations it set. The Obama administration claimed that the rate would stay below 8% if Congress passed the bill, but would reach 9% by 2010 otherwise.

Regardless of this specific instance and outcome, the question is: how should economists interpret events and data?

In physics, variables can be isolated and experiments can be controlled and repeated. So, empirical results can settle debates and advance knowledge. Physics textbooks are filled with scientists, laws, constants, experimental setups and results, and quantifiable and improving predictions.
This contrasts sharply with economics, where data usually only feeds the debate and whose textbooks contain theoretical models and historical illustrations. As much as mainstream economists claim to adhere to the empirical methodology, they actually do not proceed by experimentally falsifiable theories and a track record of predictions.

If the desired effect happens (or the undesired effects only happen later), economists and politicians celebrate.
If the expected effect doesn't happen, mainstream economists claim the baseline was wrong (things were worse than anticipated) or the intervention insufficient.
Although the first outcome seems deceptively benign and only the second usually raises questions, they both risk harmful unintended consequences and share the same pretense of knowledge.

The empirical methodology fails because of two fundamental issues: irregularity (historical events cannot be reproduced, and there are no measurable constants in man's nature) and complexity (factors cannot be controlled, isolated and often even measured).

The praxeological approach to economic theory avoids these issues because it covers a different kind of knowledge.
First, it handles the issue of irregularity by relying on logical reasoning (deduction) on the timeless axiom of action, instead of observation (induction). As I outlined in my last post, this means that theorems describe what must be, and the interpretation of empirical results plays no role in validating the theory, just like for geometry.

Second, it solves the issue of complexity in reasoning by decomposing the problem, often using thought experiments and using the ceteris paribus assumption ("all other things being the same"). For example, ceteris paribus a legislative raise of the minimum wage causes a raise in unemployment.
That isolates causal relationships by controlling and fixing all other factors.

Finally, it leaves the issue of complexity in observation outside of the realm of theory and up to the unscientific interpretation and judgment of economic historians, forecasters and entrepreneurs.

Such an economic theory yields definitive propositions and a deeper knowledge of causal relationships than physics, on the basis of praxeology and the action axiom. However it does not possess the kind of quantitative forecasting power which made physics widely successful and useful. This is disappointing and humbling, but it is not a new realization and we should be honest.
During the 20th century the neo-classical schools tried to make economics into a different kind of science than it is and their predecessors recognized.
In addition to taking economics into a detour and confusing theory, history and forecasting under guise of progress, this has unfortunately caused very real damage to societies. We will discuss some examples of harmful policies later, but not before we look at some major theorems in the next few posts.

It is time for economic science to return to a sound methodological foundation, for the good of the profession and the public.
As painful as revisiting half a century of so-called scientific progress may be, it is unscientific and intellectually dishonest not to clearly admit what we do or don't know, with what certainty and caveats, and why. Empirical results can be useful, but they cannot have the same status as economic laws or invalidate such laws.

In closing, some quotes from Austrian economist F.A. Hayek's 1974 Nobel Lecture, The Pretence of Knowledge which emphasizes this issue:

"It seems to me that this failure of the economists to guide policy more successfully is closely connected with their propensity to imitate as closely as possible the procedures of the brilliantly successful physical sciences - an attempt which in our field may lead to outright error."

"(...) and I confess that I prefer true but imperfect knowledge, even if it leaves much indetermined and unpredictable, to a pretence of exact knowledge that is likely to be false."

"To act on the belief that we possess the knowledge and the power which enable us to shape the processes of society entirely to our liking, knowledge which in fact we do not possess, is likely to make us do much harm."

Permalink | Comments (2) | September 19, 2009