WIXLIB

Page 3 of 9are needed by the international community to promote moreconservative use of antibiotics (Van Broeckel et al. 2014:1).However, as mentioned previously, if a microbe developsresistance in one region, it can quickly spread to all otherregions of the world; hence, supplementary strategies areneeded.Co-infections of human immunodeficiency virusor tuberculosis and human immunodeficiencyvirus or malariaAntibiotic resistance is accelerating because of co-infections,such as those between human immunodeficiency virus (HIV)and TB, which exacerbates the morbidity and mortality ofeach alone (Deidrich & Flynn 2011). Over a billion people areat risk from malaria - in 2016, at least 212 million cases werereported, with 445 000 fatalities, most of these being amongstAfrican children (Mandala et al. 2018). Malaria is alsodeveloping resistance to current drugs. Co-infection withHIV exacerbates the dangers of malaria (Mandala et al. 2018).Even more serious is the risk of co-infections of HIV, TB andmalaria, found to be associated with migrant workers andtravel to high-risk countriesIs antibiotic resistance a symptom of a largerproblem with the discovery system itself?Antibiotic resistance is particularly problematic becauseindustry’s ‘retreat from developing new antibiotics is leadingto a loss of expertise in both practical and theoretical aspectsof antibiotic biology’ and as ‘industry reassigns or retires itsmicrobiologists’, fewer will be trained in academia, as theknowledge base will take years to rebuild (Nathan 2004:899).However, if the problem is associated with the globalpharmaceutical R&D system itself, then it might be necessaryto supplement the system with other approaches. Academicresearch can sometimes fail to take up innovative opportunitiesto improve research productivity (Rubin & Callaghan 2019).We argue that a problem of unresponsiveness to importantsocietal research imperatives exists across the global R&Dsystem itself.As stated previously, this literature is not new (see Cowen2011; Gordon 2016), and there exists formal theoreticaleconomic models that explain why it is becoming harder tomake scientific breakthroughs (see Kortum 1997; Segerstrom1997). We suggest that if evidence exists of the lack progress todate against antibiotic resistance, then we might be headed fora ‘cliff’, whereby the resistance to learning of microbes exhaustsour last defences against them. This evidence points to animpending disaster. If disaster risk reduction is a systematicapproach to the identification, assessment and reduction of therisks of disaster, then this article seeks to identify threatsassociated with the failure of the research system to solve theproblem of antimicrobial resistance. By identifying appropriatetheory, and what it predicts about improving the effectivenessof antibiotic research, we seek to derive useful insights intohow to reduce the risk of research failure.http://www.jamba.org.zaOriginal ResearchMethodologyThis article applies a conceptual critical review methodology.In doing so, literature and theory are synthesised in order todevelop propositions. These propositions are then used todevelop a research agenda. Recommendations are thenderived from this process. The strength of this methodologicalapproach is that it provides a synthesis of novel theory andideas, and develops testable propositions. Further research isthen tasked with testing these propositions. Certain microfoundations of theory are now considered, which mightsuggest how to mitigate the threat of research failureassociated with antibiotic resistance.Theoretical micro-foundationsWhich theoretical frameworks are useful in reconceptualisingscientific research as a real-time process? What are thetheoretical mechanisms that predict how this could beperformed? Leveraging collective human intelligence isperhaps one approach to address the above questions(Callaghan 2018a).Micro-foundational theory is a body of the theory thatencompasses theoretical relationships that predict how toimprove specific aspects of the research process, so as toattain near real-time research productivity. These aspects arefed into the disaster response process, which requires realtime knowledge of how to address the problem of microbialresistance. Macro-foundational theory relates to theoreticalpredictions about the overarching context of researchproductivity, or relationships at the aggregate level. Here, welimit our discussions of the disaster response to the researchprocess, which we consider to be particularly important inlight of the failure of the research response to the problem ofantibiotic resistance to date.Building on Malone, Laubacher and Dellarocas’s (2009)ideas, Callaghan (2018a) described this body of theory as‘networked science theory’, drawing on Nielsen’s (2012)term, and Nielsen’s arguments that changes to the discoveryprocess (the scientific research process) itself herald a newage in human development. Indeed, Nielsen (2012) arguedthat those looking back from the future will identify twoperiods in human history, namely the pre-networked scienceand post-networked science eras. For practical examples ofhow the research process itself is being accelerated throughthe application of the principles of networked science, seeNielsen (2012).Nielsen’s (2012) examples offer us useful insights into thepractices of research that successfully harness humancollective intelligence. We argue here that these examplesalso offer useful insights into how to attain real-time disasterresearch response. The practice of networked science, whichwe envision to provide important guidelines for how toattain real-time research capabilities, or the potential for realtime disaster response, is shown as the top portion of thepyramid in Figure 1. The practice, however, is underpinnedOpen Access

Page 4 of 9Original ResearchReal- me disaster research responseThe permuta on challenge (Proposi on 1)The pre-permuta on mode of science will experience diminishingreturns to R&D investments.Prac ceThe knowledge aggrega on challenge (Proposi on 2)Investments in research that are focussed on improvingknowledge aggrega on will have substan ally higher payoffsthan those focussed merely on me and costs in the researchprocess.Micro-founda onal theoryBalancing societal ra os challenge (Proposi on 3)Only through a focus on balancing societal R&D ra os is it possibleto solve most contemporary societally important researchproblems.Macro-founda onal theoryFIGURE 1: Levels of theory supporting the practice of real-time disaster response.by micro-foundational and macro-foundational theoreticalframeworks.Micro-foundational theory relates to the question of how thefront line of the research can be pushed back, or how economiesof scale can be developed within the research process itself.According to Bernstein, Klein and Malone (2012:41), it is nowincreasingly possible to solve problems using networks ofhumans and computers, a phenomenon that has been termed‘collective intelligence, social computing or various otherterms’. Although computer applications and open-sourcesoftware packages have been used in drug design for manydecades, novel theory suggests that such approaches have thepotential to contribute to economies of scale in the researchprocess, if used more extensively, and if certain theoreticalprinciples are applied to this purpose. Our argument here isthat investments in developing knowledge in this area alsosuffer from the societal benefit ratio effect – the number ofpeople or resource costs affected by antibiotic resistance(the denominator) is much larger than the numbers of peopleor resources involved or invested in researching how toaddress the problem. An important contribution of ourargument is its effect as a heuristic to help to focus attention onthe problem of very small societal benefit ratios.Crowds of people can now ‘share, recombine, and refine eachother’s creative outputs’ in open idea ecologies, and theseapproaches have led to successes like youtube.com,Wikipedia.org and Linux (Bernstein et al. 2012:41). Such anapproach, although already prevalent in scientific research,might have the potential to radically reduce time and costs inthe R&D process, if used more extensively.Commercial sites like InnoCentive have demonstrated proofof concept, putting scientific problems online for the crowd tosolve (Malone et al. 2009). In this way, complex scientificproblems are solved more quickly and cheaply than ispossible with closed models of in-house R&D. The primaryconstraint of the InnoCentive model, however, is perhaps thesize of the rewards put up for the solving of a particularproblem (Callaghan 2017). Radically increasing the size of thereward might increase the incentives of solvers, andthe speed at which solutions can be found. Antibiotichttp://www.jamba.org.zaThe R&D virtualisa on challenge (Proposi on 4)The virtualisa on of the research or R&D process is a necessarycondi on to enable the research produc vity required to balancesocietal R&D ra os.R&D, research and development.FIGURE 2: Necessary conditions for solving contemporary societally importantresearch problems.resistance is perhaps a useful candidate for this approach.Micro-foundational theory therefore contributes directly toimproving the economies of scale that are attainable in theresearch process itself.Theoretical macro-foundationsThere are certain insights that have recently emerged from thefield of big data analytics. According to Kitchin (2014), theadvent of the ‘big data’ era has led to a ‘new epistemologicalparadigm’ differentiated from traditional deductive‘knowledge-driven science’. According to Callaghan (2018b),implications of this perspective are not only that interconnectivityand interdisciplinary research will increasingly be enabled byexponentially increased data availability but also that newmodes of theory development are now possible.Permutation-based scienceThe primary macro-level theoretical frameworks offered byRomer (1990) and Weitzman (1998) seem to suggest thatbreakthroughs in R&D might come through the recombinantprocesses of innovation and ensuring that ideas are shared inthe problem-solving process. Given the advent of the big dataera, however, it would seem that the ‘problem space’ of sciencehas radically expanded. If problem-solving can be representedas a topography, or a space, then Romer’s and Weitzman’stheoretical frameworks predict that the solutions to R&Dproblems like antibiotic resistance lie in the ability to theoriseand test permutations of presently existing knowledge.Thus, the first macro-theoretical principle of this article isderived here, that it is only the permutation-based eraof science that may hold the promise for achieving economiesof scale in the R&D process. The attainment of theseeconomies of scale might be a necessary condition for theOpen Access

Page 5 of 9solving of pernicious problems like antibiotic resistance.Therefore, we suggest:Proposition 1: The pre-permutation mode of science willexperience diminishing returns to R&D investments.We therefore argue here that the pre-permutation mode ofscience is simply not able to allocate the required researchresources in a way that matches the societal costs of certainproblems. This problem is arguably reflected in societal R&Dratios that suggest that vast positive externalities remainuntapped in the contemporary global R&D system.If it is only permutation-based science that can solve problemslike antibiotic resistance before its catastrophic consequences,we need to ask what dominant constraints exist to itsadoption? What overarching theoretical guidance can weglean from macro-level theory? A consideration of the seminalknowledge aggregation problem might offer us usefulinsights in this regard.The (seminal) knowledge aggregation problemProbabilistic innovation theory (PIT) (Callaghan 2017)predicts that radically improved cost and time savings canresult when problem spaces are ‘populated’ with largenumbers of solvers. This theory suggests that by balancingthe societal innovation ratio, it is possible to unfold a problemspace by populating it with large numbers of people inproportion to the importance (those affected by) of aparticular knowledge problem. A real-life example of this canbe found in the form of InnoCentive, a site that puts scientificproblems online to be solved for reward, using an open call.Probabilistic innovation theory seeks to reframe the focus oftheory development through the use of a heuristic to focusresearch problem-solving on knowledge aggregation, whichis taken to be the dominant constraint for progress in R&D.Drawing from the princsiples of PIT (Callaghan 2016a, 2016b,2017, 2018a), this can conceptually described as follows:ξ f(time * resources)ᵊThis heuristic suggests that the probability of solving anysolvable problem (ξ), although necessarily a function ofcomplex interactions of time and resources, is primarilysubject to the exponential influence of constraints associatedwith the knowledge aggregation problem ( ). This heuristicaccords with predictions of the importance of knowledgeaggregation (Hayek 1945; Von Hippel 1994) in knowledgecreation, particularly in terms of how to aggregate tacitknowledge (Nonaka 1994), which is at the level of theindividual (Polanyi 1973). An implication of this heuristic isthat a focus on improving knowledge aggregation mightensure that theory development is targeted on the mostimportant link in the aggregated problem-solving process. Onthe basis of these logics, the following proposition is derived:Proposition 2: Investments in research that are focussed onimproving knowledge aggregation will have substantiallyhigher payoffs than those focussed merely on time and costs inthe research process.http://www.jamba.org.zaOriginal ResearchIt must be noted that the quality of inputs into the Wikipediatype research process is a concern. An ethical frameworkpremised on the principles of post-normal science (Funtowicz& Ravetz 1994), however, suggests useful principles that canbe applied to maximise accountability and transparency inthe open research process.Have certain characteristics of the academic systemcontributed to a widening societal R&D ratio? From theperspective of computer science, Brookes (1996) issued thefollowing warning over two decades ago:[ W]e tend to forget our users and their real problems, climbinginto our ivory towers to dissect tractable abstractions of thoseproblems, abstractions that may have left behind the essence ofthe real problem. We talk to each other and write for each otherin ever more esoteric vocabularies, until our journals becomeinaccessible even to our society members, and publicationproperly commands a higher price from the author in page feesthan from the reader in subscription fees. (p. 62)Key to real-time knowledge aggregation is perhaps theleveraging of collective intelligence. Brookes (1996:64) arguedthat ‘intelligence amplifying systems can, at any given levelof available systems technology, beat AI [artificial intelligence]systems’, or in other words, ‘a machine and a mind can beat amind-imitating machine working by itself’. To obtaineconomies of scale in the research process, it is necessary toleverage human problem-solving capability, and this isperhaps only possible if the search for opportunities to applynovel technologies to the research process becomes animportant stream of research, in itself. In other words, ifsufficient researchers populate the numerator of its societalR&D ratio, and if the contributions of these researchers areleveraged using appropriate technologies, then this mightturn the tide in certain societally important areas of science,including the development of new antibiotics; hence, thefollowing proposition:Proposition 3: Only through a focus on balancing societal R&Dratios, is it possible to solve most contemporary societallyimportant research problems?The corollary of this proposition is that problems likeantibiotic resistance can only be consistently solved througha radical re-conceptualisation of our education systems andthe supply of researchers in the system. Knowledge of theincentives structures within science itself may also requirecareful scrutiny. If the attractiveness of laboratory-basedscience is not sufficient to attract the numbers of problemsolvers required for a balanced societal R&D ratio, then noveltheorising is required urgently.Virtualisation of the research orresearch and development processAs stated above, strengthening the numerator of the societalR&D ratio might occur by simply populating an unfoldedresearch problem space with large numbers of researchers.This can currently be performed using methodologies likecrowdsourced R&D. If technology can be used to leverage theOpen Access

Page 6 of 9contributions of these researchers included in the numerator,then their societal impact might be substantially increased,relative to the denominator, or the numbers of people whobear the consequences of a failure to solve a societallyimportant knowledge problem. This type of approach is notnew. According to Brookes (1996:64), more attention shouldbe paid to how humans can work synergistically withartificial intelligence, and computer graphics research may bekey to this.Theory development should therefore be targeted at thepotential to leverage, or amplify, collective intelligence(Malone et al. 2009; Nielsen 2012) in the knowledge creationprocess, but with a specific focus on knowledge aggregation,and targeted attempts to transcend the knowledgeaggregation threshold. Research should seek to identify andquantify the dimensions of the knowledge aggregationthreshold associated with different societally importantresearch problems (Callaghan 2017).We suggest here that appropriate processes within theresearch system should be mapped, and assessed as to theirdegree of resistance to virtualisation, with a goal ofvirtualising whichever processes and entire systems that areappropriate in the pursuit of a virtualised discovery system.A virtualised discovery system is a system of research thattaps into the near-zero marginal cost model of digitalprocesses. Although the arguments made here are broad, andtheir granularity is necessarily limited, being at a certainlevel of abstraction, we suggest that if viewed through theheuristic lens of knowledge aggregation and societal R&Dratio logics, then these insights may provide a useful focusfor further theory development. The following proposition isderived from these discussions:Proposition 4: The virtualisation of the research or R&D process isa necessary condition to enable the research productivityrequired to balance societal R&D ratios.A corollary of this proposition is that a radical reconceptualisation of the research process is necessary inorder to solve problems like antibiotic resistance. In itscurrent form, it seems as if the probability that it canadequately address threats such as antibiotic resistance isuncertain. These propositions are taken to be necessaryconditio

1School of Economic and Business Sciences, University of the Witwatersrand, Johannesburg, South Africa 2Milpark Business School, Milpark Education, Johannesburg, South Africa Corresponding author: Chris Callaghan, chris.callaghan@wits.ac.za Dates: Received: 24 Nov. 2018 Accepted: 31 Jan. 2019 Published: 12 Mar. 2020 How to cite this article: