brfid

ChatGPT, SOCs, and the Productivity Paradox

What will happen when AI sweeps through business? It will be just as profound as desktop computers or the Internet, and it will be just as insignificant for overall productivity.

Or will it

Digital computers pose significant challenges in quantifying and comparing their magnitude as capital. Capital that consists of computing devices is often (and here) called computer capital. The magnitude of computer capital can be measured with combinations of one of two basic strategies. The dominant method measures performance, with measurements that range from the simple Millions of Operations Per Second (MIPS) to complex indexes that attempt to also account for qualitatively new functions (e.g. hardware support for machine learning). Benanev (2020) explains how this method makes the production of faster and more sophisticated computers appear as if more computers are being produced. Conversely, computer capital can simply be measured by its market value, with appropri

brfid

New Research: Mission Engineering and Multi-Domain Operations

I'm serving as Co-PI for a Lockheed Martin grant on Multi-Domain Operations (MDO). That's a warfighting concept described by the US Army Training and Doctrine Command (TRADOC) as the "rapid and continuous integration of all domains of warfare." At its core, it integrates the disparate technologies, strategies, and command structures that characterize the varied warfighting domains (subsea, sea, land, air, space, cyberspace). Its conceptual origins date to TRADOC's founding in 1973, which gave rise to predecessor concepts such as AirLand Battle and Full-Spectrum Operations.

MDO requires seamless and actionable communications in contested warfighting domains. As such it poses fascinating technological challenges. Indeed, linking tactical and strategic forces with com

brfid

Talk: Against the Encrypted Network Society

In a couple of weeks I'm giving a Zoom talk for the Information Security Group, Royal Holloway, University of London. It is entitled Against the Encrypted Network Society: Rethinking the Social Basis of Cryptography. If I end up developing particularly good slides I'll post them to this gist.

Here is the abstract; below (in this gist) is the cover of On Distributed Communications XII, a report by Paul Baran that I declassifed a few years back.

Since the 1960s we have been told that new computing technologies are ushering in a new era: the Computer Revolution and Knowledge Economy (1962), Global Village and One-Dimensional Man (1964), the Third (1975) and the Fourth (2015) Industrial Revolutions(s). There was never a consensus on which kind of computational techniques were behind the change. Then in the 1990

brfid

Our Major Shift in the Security Landscape

f5 Networks is a F500 firm that does application delivery networking. They've put up a lovely blog post on my NSF network security research project:

By treating networks, security components, and operations staff as part of an interdependent system, the metrics will be able to account for factors such as outstanding security vulnerabilities, strategic and long-term planning, and constituency interests, and will provide on-the-ground SOC analysts with ways to input local knowledge into higher-up decisions. This could have the potential to ignite a major shift in the security landscape by providing a powerful new framework for real-world security assessments.

Thus far their research team has embedded an academic researcher in a separate security operations center and is analyzing

brfid

Capitalism, Computation, and Causality: A Slide From A Talk

There's a lot of water-cooler buzz about the relationship between capitalism and computation. Did early computation make capitalism possible? Did capitalism lead to the elaboration of computation? And/or, should capitalism be understood as a form of distributed computation? Is capitalism an algorithm? Am I a computer? Is the solar system is one big atom? Are pre-enlightenment forms of similitude still the basis of most humanistic reasoning?

The major problem is that neither concept is well-specified. Or rather, some academic communities don't bother specifying it at all, while others specify it quite clearly but are at odds with each other over the best specification. In a talk I gave earlier this year, I mapped out some popular definitions for each, with very rough dates for each definition's emergence:

![Defining capitalism and computation](https://gist.githubusercontent.com/brfid/f

brfid

Why Do We Call Them Packets?

Alan Turing used the term packet to refer to a fixed length (but not necessarily uniform) data structure used to order a larger body of information in order to facilitate computation in a digital computer. It is in his Computing Machinery and Intelligence paper (Mind 59/236):

He wrote the paper before he left the National Physical Laboratory in 1948, where he worked alongside Donald Davies. It was Davies, in turn, who introduced packet in computer networking to denote, well, what we call packets (but not datagrams) today. (And this is where, in computer networking lore, the story begins.) Davies recalls that "after disc

brfid