Back to Summer Programs

Wolfram Summer School

June 24–July 14, 2018
Bentley University, Waltham, MA

Alumni

Arjan Puniani

Class of 2013

Bio

Arjan Singh Puniani recently graduated with Academic Honors from the University of California, Berkeley, where he studied Chemical and Biomolecular Engineering (CBE) and Electrical Engineering and Computer Sciences (EECS). He was subsequently accepted into the graduate curriculum, where he completed all coursework for a PhD in Particle Physics with UC Berkeley's Center for Theoretical Physics, summa cum laude. Throughout his academic career, Arjan held research appointments with the top institutions in a number of eclectic fields. Arjan has performed computational psychophysics research under Physics Nobel Laureate Donald Glaser at the Helen Wills Neuroscience Institute and investigated neurosurgical therapies for the highly aggressive brain cancer, glioblastoma multiforme (GBM), at the University of California, San Francisco. Recently, Arjan focused on experimental particle physics at the Lawrence Berkeley National Laboratory, where bolometric detection methods were employed to establish the Majorana character of the neutrino. Prior to graduation, Arjan worked for the Berkeley Quantum Information and Computation Center, investigating physical implementations of quantum computers and botanical means of quantum information signal processing.

In addition to basic scientific research, Arjan has pursued opportunities in the upper echelons of finance, including Technology Investment Banking at Pacific Crest Securities and Portfolio Investment Management at Goldman Sachs. Arjan's interests include fencing (sabre), equestrian polo, weight-lifting, tennis, golf, football, and racquetball. He is an avid reader of Thomas Pynchon, immensely enjoys supporting the Miami Heat, St. Louis Rams, and Real Madrid, and has travelled to Europe, the Caribbean, India, and Dubai.

Project: On the Viability Potential of Digital Safe Schemes via Time-Lapse Cryptography

Certain countries, such as the Commonwealth Nations, publicly disclose government cabinet documents after 30 years. Securely storing those documents requires a dedicated workforce tasked with maintaining an encrypted-server infrastructure preserving the data pushed to the cloud, and guaranteed (only) by the lifetime of the sovereignty. Suppose a private individual wished to perform something similar; that is, digitally protect the integrity of a binary file on the order of decades for deferred consumption. For example, if one wished to send a confidential letter to himself 30 years in the future, what is the best way of accomplishing this?

There are several promising avenues to pursue, and all relate to circumventions to computational irreducibility (there are fundamental limits to predictability to contend with). While it is tempting to simply bury a flash drive in a traditional time-lock safe, a number of shortfalls surface. How would one guarantee the structural integrity of the physical safe from the elements, let alone protection from forced entry if the location was compromised? Memory cells, found in flash drives, cannot even guarantee indefinite information fidelity due to limitations on read-write cycles, and also require optimized environmental conditions for long-term preservation. Any other physically-based protection scheme represents a specific instantiation of the general problem described above. (One exception is simply commissioning N law firms to guarantee protection of the file on-site until a disclosure date in the future, where N is the minimum Ramsey number required to guarantee delivery).

Possible candidates include:

Distributed Computing

  • Torrents—Private file dissemination may be accomplished through a BitTorrent network. The binary, much like a torrent, would not contain the content itself; rather, it would simply be metadata that enabled participants in the system to find each other and form efficient distribution groups (swarms). Uploaded content would be broken into chunks and redundantly hashed across secure nodes (or donated computing resources, e.g., SETI@home) to protect data integrity. A built-in timing mechanism pinging Coordinated Universal Time would request multiple chunks from different computers in the swarm when desired. By decentralizing distribution, desired content would not depend on a single digital storage system's integrity.

    Issues: The "Quis custodiet ipsos custodes?" Problem. That is, who or what will securely store the metadata (torrent) itself? Changes to internet protocols, chunk encryption, server/node guarantees, etc., obfuscate viability, too.

  • Bitcoin—Part of the cryptocurrency's popular appeal is independence from a central authority, but the infrastructure may prove useful for my particular research problem. The encryption protocols operate over a P2P file-sharing network to solve the double-spending problem (failure of security protocols to transfer unique ownership of digital cash in a transaction in the event of file duplication). Similar to Bitcoin, one possible solution is to employ a distributed "blockchain," a sort of electronic ledger that relies on sequentially-created local copies of the transaction details to verify the complete transfer of t. A similar cryptographic P2P protocol may be employed to protect the binary file integrities, and when coupled with time-controlled blockchain generation, the objective is achieved.

    Issues: Lifetime expectancy of Bitcoin network may not guarantee data preservation.

Exotica

  • Astronomical Distances and Finite Speed of Light—Suppose Alpha Centauri was replaced with a giant, planar mirror oriented normal to any directed transmission from Earth. In theory, a photon encoding information (such as a radio or microwave burst) could be shot at the space mirror and returned to the sender over a period of 9.58ly. Different astronomical objects over varying distances also serve as candidate reciprocators to toggle over desired return dates. With quantum key distribution (or entanglement-dependent EPR pairs), a provably secure protocol where information gain is possible only with an introduction of a disturbance, verifying data integrity is trivial, provided the transmission encounters no obstacles.

    Issues: Low-albedo astronomical objects. Diffraction, reflection, and (Compton and back-) scattering from space debris, dust, quasars, and supernovas may compromise the efficacy of quantum cryptography by irreversibly coupling the transmission to the environment, but NKS concepts may prove useful.

    Viable solutions are impossible without applying the structured methodology introduced in A New Kind of Science. While arbitrary rules applied sequentially often yield unique results, certain variants of cellular automata allow for rules that may be applied in an arbitrary order, which ultimately drive intrinsic synchronization and guarantee a particular steady-state. Controlling cell updates by manipulating the receipt of data underlies the basic mechanics of the process, so an obvious correspondence between the "Firing Squad"-synchronized CA evolution and the desired deployment time. My project intends to explore the physical implementations of the internally-synchronizing CA for the purposes of securely storing binary files for the purposes of coordinated, future deployment—a digital timed-safe.

Favorite Four-Color, Four State Turing Machine

Rule 7039607544195917936750