About Intelligence

The INTELLIGENCE newsletter was first published in May of 1984 and monthly thereafter until December, 2017. The newsletter's original subtitle was "The Future of Computing." The subtitle was changed in 2014 to "Making the World Work," the current focus of this website.

As one of the founding editors of Omni magazine, I was responsible for editing several articles on subjects like automation, robotics and artificial intelligence (AI). After leaving the magazine, I discovered a trade newspaper, American Metal Markets, that, in 1983, listed 500 companies involved with AI, robotics and related areas.

As I was researching and making preparations to launch my AI newsletter, everyone told me the title of my publication must include the term AI or artificial intelligence. I resisted, telling my well wishers that I would name my newsletter  INTELLIGENCE but not "artificial" anything. "They'll think your newsletter is for spooks and spies, that kind of intelligence," I was told.

Over the years, many intelligence agencies of the US and other governments, took subscriptions to INTELLIGENCE. These included the FBI (Federal Bureau of Investigation), the CIA (Central Intelligence Agency), the NSA (National Security Agency). They subscribed to find out what was going on in AI, neural networks and related advanced computational technologies, the areas covered by INTELLIGENCE.

During these decades, some of the most significant trends that emerged now seem obvious in retrospect. These include, most especially, the extraordinary acceleration and perpetuation of Moore's Law observing semiconductors basically doubling every two years at steady price points. Also now clear is the brittleness of earlier instantiations of AI, like expert systems. For years, those in traditional AI, business and academe, clung to the notion that neural networks (NNs) were not the answer. And, for years, NNs were only a partial answer, with difficulty scaling except in very specialty application areas.

More recently, everyone loves NNs, machine intelligence, deep learning, machine learning and their applications abound. This is thanks to several other trends. Most prominently, as a result of Moore's Law and chip innovations, like graphical processing units (GPUs), combining with enormous data sets, many neurocomputing approaches achieve excellent results, especially in areas involving pattern recognition. But, of course, we humans, too, are pattern recognition experts.

The flow of information and ideas about areas related to NNs has gathered stunning speed in the latest years as the amount of data, observation, reporting and opinions about these computational approaches has geometrically expanded. If you need the news about NNs and other advanced technologies, the Internet is your best source and a constantly updating repository.

It has been my pleasure, my excitement and my education to edit and publish this newsletter. The above noted flow of information already does more than is needed in order to keep informed and keep up. I'll be moving on from newsletter publishing to embrace what I consider the most important issue, now a question: How do we make the world work?

Ed Rosenfeld, New York, 2017