Research

I believe that the concept of a "paper" as a unit of scientific research has come to its end. The way we present and distribute scientific knowledge no longer serves its purpose: It has become a tool of career growth, positioning and promotion. The process of verifying the correctness of the results is broken, because the incentive structure encourages overclaiming and hype over verifiable and reproducible science. Furthermore, science reproducibility suffers due to social reasons: extreme competition in the academic job market and research funding encourages scientists to stick to their work and avoid public disagreements. Simply put, you do not want to criticize the people that write you tenure support letters, review your grant applications and referee your papers.

Yesterday it was difficult to imagine an alternative: If I do not write papers how do I communicate my work to the community? Today it is difficult to imagine why anyone would write a traditional paper (other than for the reasons discussed above). It is now possible to present your research in the form of interactive, automatically verifiable artifacts with minimal effort. Analytic calculations can be supplemented with interactive visualizations, numerical checks and symbolic, machine-verifiable derivations. Experimental work is data supplemented with interactive tools: open-source tools used to analyze it plus an exact description of the experimental setup. In this form, the verifiability of theoretical results should be manifestly transparent. The reproducibility of data analysis should be manifestly transparent as well. Reproducibility of the experiments themselves will take longer, but I am certain that it is the not-too-distant future with self-driving labs. All this does not imply that every verifiable result is correct, but it will ensure that the road from assumptions to conclusions is clear.

Here I will post my explorations of the format in which scientific results can be presented. In my career I wrote around 60 "papers" and, in my spare time, I will attempt to present some of them in this new format. As my vision evolves, I will improve the artifacts.

P.S. I have carefully hand-selected every token on this page as well as each research page.