ROUGH DRAFT authorea.com/7565
Main Data History
Export
Show Index Toggle 0 comments
  •  Quick Edit
  • Mass Effect Disruptive Innovations in Research Science

    Big Science

    En masse, science is becoming really big science. An new paradigm for science is emerging that is making collaboration in science much easier. There are some very strong advocates for open science that aim to evangelise the science community at large. At the same time, Software as a Service (SaaS) tools are enabling a more open science. Cloud based storage and computing services provided by Google, Dropbox, and Amazon are pervasive in research science. Software is making it easier for researchers to communicate, share, and access information. Software is enabling the Big Data infrastructure that industry demands and it is beginning to support the scientific community, too. But, the pantheons of science are fighting back with more technologies, more variety, more data formats, more ideas assumptions which complicate software development. There is a big difference between research science and software development, but big science hinges on a smooth interface between the dissemination of research and software requirements.

    The tension in science is being driven by extremely demanding design spaces. The scale of scientific exploration has greatly exceeded reductionist approaches. Science has been proliferated fringe communities of domain experts who, with experience, have the experience and rite to guide their field, hell someone had to. These domains have unique lexicons, lots of acronyms, and enough secrets to imagine there must be a secret handshake. The factions in science are being required to work together and its real awkward. We talk funny, we talk pidgin. Science policy is addressing the digital aspects of the language, or information barrier, but the domain lexicons are a troublesome bit of language. The language of science is going to evolve over the coming years as we witness some disruptions of globalized science.

    Globalized science is going to confront an ever presented cultural gap

    There is a distinct generation gap in science. The current thought leaders from GenY and prior have grown up on the fringe of science, they have worked in small fishbowls, small communities that are indicative of the times they were raised in. The newest breed of scientist has learned to exist is a mobile and social manner shattering all ideas of community boundaries. For the new generation, science isn't designed for them; they lead disparate digital lives in science and their personal life. This dissonance will not exist for long, really their is no point in resisting.

    Before our eyes science is becoming more open and it is going under the radar of many scientists. The driving force behind this transition is software; software that the mobile and social generation are building. I really hope that in the future that no scientist ever scientist says, "\textit{I'm going to build a GUI for it.}". The early open science evangelists are using software that is designed for science like Authorea, Plot.ly, and Figshare . Larger infrastructures like Google, Dropbox, and Amazon Web Services are offering semi-open science tools that scientists use often. The common element behind each piece of software is that it is all Software as a Service; the software is mobile and social.

    The current pressures on research science and heavy

    Isolated debates are occuring on the current state of scientific data, scientific codes, peer review publication, and STEM topics. Individual organizations are attempting reductionist models to solve a truly emergent problem. There are four synergistic and sometimes conflicting products of research science.

    • Data Management
    • Code/Software Management
    • Dissemination of Research Results
    • Teaching the Next Generation Workforce

    Each component of the scientific research process influences the other, but yet there are serialized efforts to support each component individually. These efforts are fighting uphill battles, the masses will sort this out. Embrace it and focus energy in better places.

    Data archiving and code/software management are topics that the scientific research community is addressing to serve their specific needs, but they are building non-scalable, non-transferable ideas. Data and Code Management are topics that the software community is addressing.

    The dissemination of research results is truly marring the progress of science. Researchers are stumped, progress isn't moving forward as science has promised. The definition of science is going to change. Historically, science has been driven by isolated and successful collaboration efforts, but the new social and mobile generation is begining to redefine what collaboration means. Science is getting bigger.

    Researchers are concerned with the quality of figures, tightness of sentences, and word counts when trying to express their positive contributions to the knowledge base of science. This has largely become the responsibility of science; researchers disseminate their best and proudest results. Cool, thanks guys, you are awesome. But what about the process! I don't give a shit about your product. Science is process driven. Scientists are artists. They work in a limited design space of experimental tools, statistical tools, and computational simulations; they drive thier experimental process off of assumptions and intuition; science can never be proven to be right; stop trying to do that. What science can do is stop making shit up, they will do science, they will record the process better, they will not put the cart before the horse.

    The Venn diagram in Fig xx. illustrates each scientific product. New collaboration models are being overlain on top existing software architectures. Github and collaborative Latex (e.g. ShareTex, Bibtex) editors are being merged to cite data and code in scientific publications. Cooperative efforts are stream lining the scientific process to get from vetted scientific research to the final publication.

    The problem with the current paradigm of science is that it is still steeped in tradition that is inhibiting its ability to its best. Most scientific progress is being evaluated under a traditional accounting model; progress is vetted by small groups of peers that are domain experts in a subject area; this proliferates as strong bias from traditional scientific practices. In "The Lean Startup", Eric Reis defines a startup as "a human institution designed to create a new product or service under conditions of extreme uncertainty". His Lean Startup philosophy, that is adopted from lean manufacturing ideas, has driven many successful startup businesses; the same ideas have infiltrated divisions of enomorous corporations too. The future has become too uncertain, we are building different ideas, not better ones. When the ideas are different, they cannot be evaluated on that past. Reis brings forth the idea of innovation accounting to value progress and learning rather than improvement. To me, science has been the bleeding edge of uncertainty, the scientific world was easier to navigate. Now there is a gigantic generation gap between the agile, lean software development world that is placing demands on science whos progress is still waterfall development with a strong focus and less the customer. Science is a public service, it is a responsibility to satisfy the demands of society before our intellectual curiousities.

    The future doesn't need to be better. It needs to be different. I want nostalgia forever, but you can now talk to someone on the other side of the world and see their face in real fucking time.

    The boundaries of scientific disciplines are rapidly becoming transparent, science will need to be uniform because wildly complex design spaces to build multiphysics technologies, science will become democratized.

    Loosen your belt Science you ate too much

    Science policy is trying to rework the boundaries of science, to contain their boundaries in nice ontologies that haplessly describe science. These efforts remain futile because science is evolving. Everyone has ten pounds of shit in a couple of five pound sacks and everyone thinks everyone else's shit stinks worse. A growth movement needs to happen where new roles are defined. It is clearly happening that we are in the age of data science where anyone who wrote a linear program is one. The data scientist, the domain agnostic number cruncher big data mofo, is going to be interfacing very heavily with domain expert data scientists. This interface is critical in aligning the physical sciences with emerging technologies.