Quantum BS: piling it higher

Hans Bachor recently gave a talk at UQ, Hype and Trust in Quantum Technologies

Trust is a core value in science, trust in data, analysis, concepts, models. This is achieved in physics by open publishing, scientific discourse, testing, repeating experiments, asking critical questions and designing new tests. Fortunately, science is self-correcting in the long term. Hype includes predictions which sensationalise scientific discoveries and exaggerate the future impact. Increasing competition for funding, visibility or job security can make this more attractive. But it also erodes trust in science by the public and investors and has negative social effects on us the researchers. How can we balance them?

I think this problem more broadly reflects the way universities have become to imitate the social context they are imbedded in, rather than being a critique of those societies.

The sociologist Christian Smith eloquently described the emergence of BS in universities, several years ago.

Emergence talk at University of the Philippines

Today I am giving a seminar on emergence at the University of Philippines Los Banos, in the Institute of Mathematical Sciences and Physics.

Here are my slides.

10 key ideas about emergence

Consider a system comprised of many interacting components. 

1. Many different definitions of emergence have been given. I take the defining characteristic of an emergent property of the system is novelty, i.e, the individual components of the system do not have this property.

2. Many other characteristics have been associated with emergence, such as universality, unpredictability, irreducibility, diversity, self-organisation, discontinuities, and singularities. However, it has not been established whether these characteristics are necessary or sufficient for novelty.

3. Emergent properties are ubiquitous across scientific disciplines from physics to biology to sociology to computer science. Emergence is central to many of the biggest scientific challenges today and some of the greatest societal problems.

4. Reality is stratified. A key concept is that of strata or hierarchies. At each level or stratum,  there is a distinct ontology (properties, phenomena, processes, entities, and effective interactions) and epistemology (theories, concepts, models, and methods). This stratification of reality leads to semi-autonomous scientific disciplines and sub-disciplines.

5. A common challenge is understanding the relationship between emergent properties observed at the macroscopic scale (whether in societies or in solids) and what is known about the microscopic scale: the components (whether individual humans or atoms) and their interactions. Often a key (but profound) insight is identifying an emergent mesoscopic scale (i.e., a scale intermediate between the macro- and micro- scales) at which new entities emerge and interact with one another weakly.

6. A key theoretical method is the development and study of effective theories and toy models. Effective theories can describe phenomena at the mesoscopic scale and be used to bridge the microscopic and macroscopic scales. Toy models involve just a few degrees of freedom, interactions, and parameters. Toy models are amenable to analytical and computational analysis and may reveal the minimal requirements for an emergent property to occur. The Ising model is a toy model that elucidates critical phenomena and key characteristics of emergence.

7. Condensed matter physics elucidates many of the key features and challenges of emergence. Unlike brains and economies, condensed states of matter are simple enough to be amenable to detailed and definitive analysis but complex enough to exhibit rich and diverse emergent phenomena.

8. The ideas above about emergence matter for scientific strategy in terms of choosing methodologies, setting priorities, and allocating resources.

9. An emergent perspective that does not privilege the parts or the whole can address contentious issues and fashions in the humanities and social sciences, particularly around structuralism.

10. Emergence is also at the heart of issues in philosophy including the nature of consciousness, truth, reality, and the sciences.

The interplay of ecological and evolutionary dynamics: immigration, extinction, and chaos (and DMFT?)

"Ecological and evolutionary dynamics are intrinsically entwined. On short timescales, ecological interactions determine the fate and impact of new mutants, while on longer timescales evolution shapes the entire community."

 Spatiotemporal ecological chaos enables gradual evolutionary diversification without niches or tradeoffs       Aditya Mahadevan, Michael T Pearce, and Daniel S Fisher

Understanding this interplay is "one of the biggest open problems in evolution and ecology."

New experimental techniques for measuring the properties of large microbial ecosystems have stimulated significant theoretical work, including from some with a background in theoretical condensed matter physics. For an excellent accessible introduction see:

Understanding chaos and diversity in complex ecosystems – insights from statistical physics

This is a nice 2.5-page article by Pankaj Mehta at the Journal Club for Condensed Matter. He clearly introduces an important problem in theoretical ecology and evolution and describes how some recent work has provided new insights using techniques adapted from Dynamical Mean-Field Theory, which was originally developed to describe strongly correlated electron systems. 

Here are just a few highlights of the article. It may be better to just read the actual article.

Fifty years ago, Robert May "argued that the more diverse an ecosystem is (roughly defined as the number of species present), the less stable it becomes." He derived this counter-intuitive result using a simple model and results from Random Matrix Theory. This is an example of an emergent property: qualitative difference occurs as a system of interacting parts becomes sufficiently large.

"One major deficiency of May’s argument is that it does not allow for the possibility that complex ecosystems can self organize through immigration and extinction. The simplest model that contains all these processes is the Generalized [to many species] Lotka-Volterra model (GLV)".

"Despite its simplicity, this equation holds many surprises, especially when the number of species is large".

Another case of how simple models can exhibit complex behaviour.

One special case is when the interactions are reciprocal – how species i affects species j is identical to how species j affects species I. "In the presence of non-reciprocity the system can exhibit complex dynamical behavior including chaos." Understanding this case was an open problem until the two papers reviewed by Mehta. For a detailed but pedagogical introduction see:

Les Houches Lectures on Community Ecology: From Niche Theory to Statistical Mechanics, Wenping Cui, Robert Marsland III, Pankaj Mehta

This is relevant to understanding the origin of the fine grained diversity observed in sequencing experiments of microbial ecosystems.

Aside: de Pirey and Bunin "derive analytic expressions for the steady-state abundance distribution and an analogue of the fluctuation-dissipation theorem for chaotic dynamics relating static and dynamics correlation functions."

"Using a DMFT solution, they derive a number of remarkable predictions... in the chaotic system the species fall into two groups: species at high abundances and species at low abundances near the immigration floor. de Pirey and Bunin show that even in the chaotic regime, the number of high abundance species in the ecosystem will always be less than the May stability bound. This result is quite surprising since it suggests that ecosystems self-organize in such a way that the high abundance species still follow May’s diversity bound even when they are chaotic."

Should Ph.D. students choose to teach?

In Australia, most Ph.D. students are fully funded by scholarships to allow them to focus on their research. This is unlike in the USA where many students must be TAs (teaching assistants) to be paid. 

In most Australian universities, such Ph.D. students can earn extra income by being tutors (same as TAs) for undergraduate courses. Many do, as earning extra money is attractive. Ph.D. students doing teaching saves universities tons of money as it means they don't need to hire and pay permanent academic staff to do this tutoring.

What is my advice to students who have this option? Here are some of the advantages and disadvantages for a Ph.D. student doing such tutoring.

Advantages

You earn additional income.

Having teaching experience listed on your CV may help you get a faculty position at some institutions. For example, in Australia, this seems to be almost a pre-requisite these days. Furthermore, if you can be innovative, and get high student evaluations, that may be viewed favourably. But that really concerns lecturing and not tutoring.

You usually learn a lot from teaching, even lower-level courses.

It can be enjoyable and satisfying. It can provide a break from thinking just about your Ph.D. research.

If you are fortunate enough to eventually get a faculty position this experience will make it easier to handle the formidable challenges of starting out teaching.

It may create some goodwill towards you in your department. You may be seen as a team player and a good departmental citizen.

You may need the money. For example, if you are supporting a family or if you are from a Majority World country and want to send money home to extended family.

Disadvantages

Foremost, it can consume a large amount of time and energy that reduces your research productivity. It reduces your mental space. You may lose research momentum and not complete your Ph.D. on time.

There can be a significant financial opportunity cost. Suppose that doing the teaching delays you completing your Ph.D. by six months. The lost six-month salary will probably be much greater than the amount you earned from doing the teaching.

It can be frustrating to deal with students who are not that interested in learning and are only concerned with grades. Furthermore, there may be the added stress of having to deal with students who make formal complaints about your teaching or their grades.

It may not add a lot to your CV, particularly if your student evaluations are average. They will probably be average or even below average since you are starting out.

If you don't do a stellar job and/or there are a few disgruntled students your reputation in the department may suffer, perhaps unjustly.

On balance, I think it depends on the individual: their personal financial situation, personality, career goals and stage in the Ph.D. In some cases, I encourage people to do this, although only for one or two semesters. In other cases, I discourage it. The main thing is to make a well-informed decision which takes into account the pros and cons. 

Students also need to be wary of the vested interests of faculty and university management that will push them towards teaching, possibly against the student's best long-term interests.

Aside. I often forget what posts I have written in the past. I really thought I had written this post before. All I could find is one on Should postdocs teach?

I welcome comments, particularly from current and former Ph.D. students who have negotiated this issue. What would you advise?