You are here

Superquestions for Supercomputers

Avoiding terra flops from misguided dependence on teraflops? (Part #1)


Introduction
Precedents and complementary initiatives
Reservations
Cognitive implications?
quantity?
Potentially relevant superquestions
Clustering superquestions in terms of WH-questions
Refining and reconfiguring the set of superquestions
Possibilities of artificial intelligence
Conclusions
References

[Parts: Next | Last | All ] [Links: To-K | From-K | From-Kx | Refs ]


Introduction

This is a reflection on the press release announcing that Supercomputers ramp up to tackle global societal problems (Science and Technology Facilities Council, 17 November 2010). It argues that supercomputers of the future, capable of rapidly crunching vast amounts of data way beyond the existing capabilities of current technology, will spearhead the development of new drugs, new sources of energy and environmental monitoring. It indicates that:

  • the Fujitsu Open Petascale Libraries  project (OPL) is a global collaborative initiative created to develop a publicly available mathematical library that will facilitate the development of the software required to run on next-generation petascale supercomputers, capable of performing quadrillions of calculations per second. These are typically measured technically in terms of FLOPS (an abbreviation for the computing term floating point operations per second, teraflops being 1012 flops, and petaflops being 1015).
  • petascale computers are capable of quickly performing large-scale and advanced computations that cannot be solved using normal computers. As such, they are vital tools for solving the important issues facing society, including improved healthcare, the development of new medicines, materials, strategies for environmental disaster prevention as well as for basic scientific research including the origin of matter and the history of the universe.

In claiming to focus on "global societal problems" and the "important issues facing society", the initiative follows in a long tradition of approaches to global simulation and world modelling (Balaton Group, Society for Modeling and Simulation International, Sentient World Simulation, Joint Simulation System,  and the European FuturIcT project). These all raise questions like who defines what are the "global societal problems" and the "important issues facing society" -- and to the satisfaction of whom? How is such definition achieved? Which problems are included in any such definition -- or excluded (possibly without consideration)? When is such definition undertaken -- and where and why?

The concern here is the nature of the questions for which such computing power is being developed, as well as the manner in which questions some would consider relevant to "global societal problems" are likely to be omitted from such exploration. The point is well made by the example of a particular dimension from climate change considerations by the IPCC in its Fourth Assessment Report (H.-H Rogner, et al., Introduction. In Climate Change 2007: Mitigation. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change) it is stated that:

The Kaya identity (Kaya, 1990) is a decomposition that expresses the level of energy related CO2 emissions as the product of four indicators: (1) carbon intensity (CO2 emissions per unit of total primary energy supply (TPES)), (2) energy intensity (TPES per unit of GDP), (3) gross domestic product per capita (GDP/cap) and (4) population....

The challenge - an absolute reduction of global GHG emissions - is daunting. It presupposes a reduction of energy and carbon intensities at a faster rate than income and population growth taken together. Admittedly, there are many possible combinations of the four Kaya identity components, but with the scope and legitimacy of population control subject to ongoing debate, the remaining two technology-oriented factors, energy and carbon intensities, have to bear the main burden.... [emphasis added]

No further reference is made to this factor by IPCC.

The concern here is the nature of the "superquestions" meriting answers -- in the light of the engagement of those who would want them answered, and who may well be called upon as tax payers to finance the development of supercomputers and their use.

It follows from earlier interest in questions and what may be readily neglected (Generating a Million Questions from UIA Databases: Problems, Strategies, Values, 2006; Global Strategic Implications of the Unsaid: from myth-making towards a wisdom society, 2003; Unknown Undoing: challenge of incomprehensibility of systemic neglect, 2008). The challenge with respect to governance has been previously summarized in a somewhat analogous checklist (Governing Civilization through Civilizing Governance: global challenge for a turbulent future, 2008) to which the "superquestions" highlighted here are tentatively linked.


[Parts: Next | Last | All ] [Links: To-K | From-K | From-Kx | Refs ]