My blog has moved!

You should be automatically redirected in 4 seconds. If not, visit
http://chrisjhorn.wordpress.com
and update your bookmarks.

Wednesday 15 April 2009

Computer History Museum - Capturing our Innovation

Computers are designed, at least at the high level, by people. Computers thus in turn reflect human intellect - innovation but also copying, sheer brilliance but also blind shortcomings. Like any engineering discipline, there is much to be learnt not just about technology but also about psychology and sociology by examining the design of computers.

If you are at all interested in computing, you really should visit the Computer History Museum in Mountain View California.

The museum is hosted in what used to be a Silicon Graphics building, just off route 101, and in fact in which I recall I had several business meetings a decade ago. Now in the foyer is one of two magnificent working reproductions of Babbage's Difference Engine - the other is in the Science Museum, London. Every hour or so, two curators give an enthralling demonstration of how the machine operates, the mechanics and the arithmetic behind it, to calculate values of arbitrary polynomial equations.

But the main part of the museum is a warehouse at the back. I experienced sheer exuberance and absolute awe by the wealth of history laid out in row after row after row. The Z3 of Konrad Zuse - a German civil engineer - is there, the world's first programmable electromagnetic computer built in 1941 at the height of the second world war using 2,400 telephone relays. The 40 rack, each 8' high, 18,000 vacuum tube ENIAC from 1946, which had to be physically reconfigured for each program "load". Gene Amdhal's PhD project, the 1955 WISC - complete with bullet holes (don't ask!). The SAGE real time air defence system from 1955, with the first ever graphical user interface. The bet-your-company IBM 360 - the first machine I used. The NASA Apollo guidance computer which Neil Armstrong had to override just before the Tranquility touchdown in 1969. The wonderful Dec PDP-8, and then -11, upon which I spent many a happy hour. The first 10MFLOP machine, the CDC6600. The IMP, the original backbone of the predecessor of the internet, and on which packet-switching was first built. The first ever computer for the kitchen - the H316 - a wonderfully funny folly. The first hobbyist machine Altair 8800 for which Bill Gates and Paul Allen produced Altair BASIC. Seymour Cray's extraordinary and aesthetic CRAY-1, -2 and -3: using cable as delay lines! The actual very first Apple-1, in a wooden box, built by Steve Wozniak in 1976. They're all there: mercury delay lines, core memories, vacuum tubes, first silicon transistors, first silicon chips, first use of gallium arsenide, huge magnetic disks, first robot arms, first computer mice, it is just so extraordinary, inspiring, and provocative.

I was recently interviewed for a podcast about my huge enjoyment of Tracy Kidder's "The Soul of a New Machine" - IMHO one of the very great books of the computer industry, and written by a lay reader for ordinary mere mortals. I was therefore delighted to see a Data General Eclipse and Digital VAX 11/780 amongst the exhibits, no doubt carefully placed to glare at each other across the aisle.

Sadly though, the museum feels, to me at least, like a cemetery. The entire warehouse has a distinctive smell of stale electronics, and all the exhibits are morbid artifacts, frozen and lifeless in their dusty racks and dull cabinets. Each of them yearns to be powered up just one more time, strains to feel the surge of bits pulsing through its accumulators, registers, de-multiplexers, decoders, caches, and yes switches and lamps.

On my visit, I stayed well after the curator had announced closing time, and managed to corner him. Surely each of these wonderful machines could be powered up again ? Even if the electronics in some cases were now no longer safe and thus a fire risk, why not simulate their operation by judiciously hiding a microprocessor in the background somewhere to cycle through lamp and circuit sequences and bring some of the thrill of these machines back again ? He responded that although this had been considered, there are apparently strongly held views by the trustees that such sorcery would detract from the raw stature of the exhibits, and perhaps demean their creators. In a way perhaps, I do understand: these magnificent wonders and follies of the most innovative and creative industry mankind has ever known now lay sadly but peacefully at rest for ever.

One of the core tenets of the Science Gallery, which I currently chair, is to bring science and technology to life by having researchers and innovators in the flesh present to discuss and explain their work to curious members of the public. A science museum is interesting, and it can be fun to try out experiments yourself. But the Gallery is different: the scientists and engineers are there in front of you, and in many cases can do the experiments alongside with (or even on!) you.

It would be wonderful to have all those inspirational architects and engineers behind the incredible machines in the Computer History Museum rightfully and deservedly proudly present when one visits. Sadly I guess, this is unlikely to ever happen. Perhaps at best there can be video interviews available to try and explain the sheer excitement behind the innovations they each made. Each one of them a master of the particular universe of virtual reality that they had uniquely created.

Whatever about trying to explain the creativity and discoveries made by computer engineers to the public, how on earth would one do it for software ? Yes, I am aware of the interest group on software at the Computer History Museum itself. But could there ever be a Software History Museum, what would it inspirationally show and how would it operate ?

Its perhaps conceivable to consider a Museum, or Gallery, of the history of computer games. Perhaps also of the evolution of graphical user interfaces. But a history, an exhibit, of software in general ?

I think the real story to be told about the history of software is not about programming languages themselves - Fortran, Simula-67, Forth, Lisp, C, Basic, Prolog, Perl, Ruby and so on. Rather it is about the evolution of programming itself: the development of data structures, the unfolding of functional as well as declarative thinking, the interplay between data and logic, introspection, aspect orientation, and so forth. Very best of all, I strongly believe that this story needs to be told plainly and simply so that the world at large can understand and appreciate how software has been evolved as we learn to reflect and begin to understand how we ourselves think and reason and learn.

I can recall after commons (the formal evening meal at my alma mater) having a heated conversation with a Professor of Genetics and various humanities students over coffee. I was the sole technologist. I fervently explained that just as there can be incredible beauty in literature and music, there too can also be extraordinary beauty in the structure of software, these artifacts of the soul and mind. However, it took a trained intellect to understand this. I was of course ridiculed, and left feeling battered and appropriately flamed. But perhaps one day, some truly gifted people will produce a software appreciation, through which mere mortals can too see the wonder, exuberance and awe of great software creations.

Monday 6 April 2009

Social Newsworking - reflections on the future of newspapers

A couple of weeks ago on the last Saturday of March, the Irish Times, a popular broadsheet here in Ireland, included a full paper reproduction of the 4 pages of its very first edition, from 150 years ago and launched on the 28th March 1859. It is a fascinating read, not least because the entire front page is given over to advertisements. I also found the style of writing captivating - archaic but nevertheless highly articulate. The editorial contains the opening paragraph:


"The appearance of another Journal in the field on Irish politics demands a brief explanation from its projectors. It will be asked - Where is the room for this new competitor for public favour ? Is there any definite set of opinions which is not already represented in the Irish Press ? From the silly Radicalism of the Phoenix Club to the dense Toryism of the Orange lodge, what doctrine lacks an exponent, what party an organ ? If no deficiency be felt, is the multiplication of newspapers a good, per se, that the proprietors of the Irish Times think fit to add one voice more the existing Babel ?"


The editor, Laurence Knox - who personally signed the inaugural edition - goes on to justify the appearance of the new newspaper as a voice independent of politics or partisismship. Reminiscent of Deng's famous assertion a century later - "No matter if it is a white cat or a black cat; as long as it can catch mice, it is a good cat" - Knox goes on to assert "Every year sees a larger and larger proportion of our population indifferent to the manoeuvres of faction...They are anxious for good government, but care little in whose hands the government may be placed".

The contemporary Irish Times commentator Fintan O'Toole recently wrote an excellent assessment of the then national context in which the Irish Times was launched. He believes that Knox's aspirations were perhaps a heroic denial of the political climate, but that from the lengthy perspective of 150 years later, Knox was perhaps not as naive as first imagined.

Ironically the same issue of the Irish Times (on 28th March 2009) also contained a topical assessment of the newspaper industry, from the paper's America based correspondent Denis Staunton. He discussed Maryland's Senator Ben Cardin's initiative to use tax policy to try and save what is left of the US newspaper industry. Cardin is of the view that it is in the US national interest, for good governance and for democracy, that at least some newspapers survive.

Devoting the entire front page to advertising may have worked well in 1859, but with advertising revenue now under immense pressure and readers migrating to free global news sources on the internet, something has to change. Web based newspaper subscriptions in general have not been particularly successful. Web based advertising is a challenging model for newspapers, in the face of Google.

What can be the future of the Irish Times, and newspapers like it ? Perhaps tongue in cheek, last January I responded to a request from the Irish Times for some crystal ball gazing - what was the company to watch in 2009 - by identifying the Irish Times itself! Will the Irish Times, and newspapers worldwide like it, survive in its current form by January 2010 ?

In my view, there are three major aspects of the essential content of a newspaper: reporting, investigation, and commentary.

The role of reporting news and events would seem well fulfilled by the density of the web itself. Timely notification of events can be reported almost by anyone: the use of Twitter by Jim Hanran, an ordinary member of the public, to first report on the crash landing of US Airways 1549 in the Hudson is an obvious example. What therefore does the traditional news reporter bring to the web ? The general answer should surely be honesty: readers are generally attracted to reputable reporting, despite occasional reporters such as The News York Times' Jayson Blair. Honesty on the web is always difficult to certify, but other web based systems leverage reputable participation - eBay auctions come to mind, perhaps also Wikipedia's consensus mechanism. However honest and reputable reporting on the web certainly is not the sole domain of newspaper reporters.

What about investigative reporting ? Most democracies can point to occasions when an investigative reporter brought malpractice to light. The Washington Post reporters Bob Woodward and Carl Bernstein's exposure of the Watergate scandal is possiby the most renowned. Here in Ireland we still mourn the Sunday Independent's Veronica Guerin, assassinated for her investigative reporting of drug trafficking.

Investigative reporting surely serves the public good. Will our democracies come under threat if there is no longer a business model for newspapers to exploit, and thus no money for investigative reporting ? This thinking seems to overlook the integrity of most members of the public: ethical employees can be appalled by what they see in their organisations, and can and do speak out. Sherron Watkins at Enron blew the whistle on her employer's malpractice. In Ireland, James Gogarty's whistleblowing led to the Mahon tribunal to investigate payments to corrupt politicians. With the internet, whistleblowers can quickly bring concerns to global public attention - India now has several public web sites (eg corruption in india, and whistleblowers) devoted to whistleblowing, particularly following the murder of Sri Satyendra Dubey. Democracy and accountability is probably best supported by well considered whistleblowing processes, which both protect a valid whistleblower but also protect an organisation from inaccurate allegations. Democracy need not rely on the finances of newspapers to underwrite investigative reporters.

Commentary, opinion and analysis is perhaps where great newspapers add most value, separate the significant from the trivial, and correlate patterns to extract observations. But there appears no inherent reason why articulate observation should just be the preserve of a newspaper: the online community and bloggers can equally serve society at large.

Clay Shirky recently blogged that the current business model of newspapers seems doomed, and replacement models have yet - if ever - to appear. Jeff Jarvis argued back in 2005 that saving newspapers is not the same as saving newspaper jobs. Perhaps it really is time to confront the unthinkable: the newspaper, as we know it, may be doomed.

Does the internet really signify the demise of the newspaper ?

Most newspapers do serve the public good. Senator Ben Cardin's Newspaper Revitalisation Act would enable US newspapers, if they so chose, to become not-for-profit organisations, thus benefitting from tax exempt advertising and subscription revenue. Members of the public could make tax deductible contributions to their favourite newspapers. I understand that public service broadcasting is similarly tax exempt in the US. Using the tax code in this way may be an approach, but I do wonder how effective newspapers may be able to operate if they come to rely on regular voluntary donations from their reading public to cover their costs of distribution and of professional staffing.

Are there any other different business models for those online newspapers which do serve the public good ? In Ireland, we have a national license which every TV set owner is supposed to pay to the Government, which in turn is used to support public service broadcasting. Could there be an online tax for public service newspapers ? Perhaps a Government could impose a tax on all internet service providers nationally, since they represent the national access points by the public to the internet: then directly, or via a neutral third party agency - use the money collected to allocate to online public service newspapers. An interesting thought perhaps, but I'm unsure the economics would really work: would internet users really accept what would amount to a personal daily tax of several euros just to support online newspapers ?

Some have given thought to a micropayment model for individual newspaper contributions. This would presumably operate something akin to iTunes: individual newspaper articles could be disaggregated - like tracks from a music CD - and made individually readable and paid for. But a musical track is a marketing teaser to promote itself: if you hear a piece of music and like it, you'll want to hear it again (and perhaps again and again) and so may be tempted to buy it from an online store. A newspaper article is typically by contrast a singular experience: having read it, you are unlikely in most cases to seek to read that particular article again.

As I reflected on the demise of the newspaper, I also reminded myself that a newspaper is a mirror held up to reflect society and the world to the eyes of its readership. It informs its readers about what is happening, and tells them something about themselves and their community, in the terms and ways to which they relate. If newspapers were to disappear due to the impact of the internet on their business model, what could be done on the internet to replace these lost mirrors on society and the world ?

In a newspaper, a selected few contribute articles and news, with feedback from ordinary members of the public filtered through the letters to the editor page. In the internet world, everyone can openly contribute. We can read posts and comments and blogs and emails and tweets from anyone. We can use Google to try and find interesting items, but even then it is easy to get lost in the morass, and perhaps miss what we ourselves find particularly interesting and significant. The specific newspapers we buy by contrast in general do a good job in only presenting articles which we ourselves find interesting and significant, filtering out the mundane, and presenting information and comment which in general match our own particular ethos and values. Any online replacement of a newspaper is going to have to do something similar.

In my view, wiki technology already provides the basis. A wiki is a web site in which not only can anyone read its content, but anyone can also edit the content, as result of Ward Cunningham's work in the 1990. At first this sounds extraordinary - what prevents somebody from editing accurate content and defacing it ? In fact, the answer is nothing, and defacement happens - however, the wiki keeps a record of who edited what, and what was the previous version (and the previous version to that and so on). Anarchists and political pundits can quickly be identified by the community of like-minded readers, isolated from the wiki so that the wiki refuses to accept any further edits from these rogues, and the maverick changes made by these fraudsters are quickly unwound.

The consequence is that the community of bona-fide contributors work together to make the wiki better and better, capturing the best inputs from everyone, and discarding the weaker edits. The system becomes Darwinian - in the view of a specific community, the best survives for that community, and the weakest is dropped. A wiki thus captures the "wisdom of a crowd" - the collective wisdom of a particular community of readers and contributors.

Social networking sites - such as Facebook, LinkedIn and others facilitate loosely coupled communities to evolve. Communities of people with common interests can quickly emerge.

Is it not possible to think of online newspapers evolving to become social newsworking ? A (potentially large) group of people with common interests and ethos could share their reflections on society and the world at large using technology derived from the current generation of wiki support and of social networking. Different social newsworks would of course emerge, reflecting the diverse opinions across societies. Each social newswork would combine - as today's paper based newspapers do - reporting of news and events, including of an investigative nature, with analysis and comment. But unlike a newspaper of today, a social newswork can reflect the wisdom of an entire community of like-minded individuals, rather than just of its editor, its staff and its contributors to its letters pages.

Social newsworking as a concept is not novel. Apart from user contributed newsites such as Reddit and Digg, one of the most interesting from the my perspective is Newsvine. Anyone can write articles, and discuss items submitted by anyone, including by professional journalists, but unlike a wiki cannot update in place previously submitted text. But in my view and as I understand it, Newsvine just aggregates and links articles together, but does not merge and synthesise. Some newspapers - such as The Washington Post and The New York Times - also actively link to internet articles outside of their own web site, and thus attempt to aggregate: but there is no merge and synthesis.

Wikinews is another social newsworking site, and does allow users to edit and update the text of others, just as for Wikipedia encyclopedia entries. However Wikinews itself aspires to offer a neutral perspective, and to be even handed.

My own view is that social newsworking can only succeed if each site in fact does not try to be even handed and neutral, but instead biased towards the ethos and philosophy of its community! Misinformation and sensationalism arguably thrive on sites such as Digg and Reddit because anyone can post, and anyone or small group can contrive to vote specific postings up or down the popularity charts. On the other hand, wiki technology could be used by a group of like minded individuals to ensure that only postings which match their ethos survive the evolutionary editing process. Rather than trying to be all things to all people and thus neutral - such as wikinews - it is probably more pragmatic to instead focus a social newsworking site around a community of like minded people who can reinforce their particular beliefs. After all, in my view, that's what our current newspapers already try to do.

A social newsworking site can of course be a free service to its community. However, an online community of like minded souls is more likely to be a successful opportunity for targeted advertisements, which raises the possibility of an effective and focussed business model. Perhaps the advertisements themselves could even become wiki based, and thus allow the content of an advertisement itself to be updated and edited by the community of its consumers. A social newsworking site could, in my view, have a viable business model if so desired, despite competition from majors such as Google or even Craigslist.

Laurence Knox, the first editor of the Irish Times, saw in 1859 - perhaps pretentiously - that his newspaper would represent a particular section of Irish society, distinguishing itself from others sets of opinions already available in the Irish press, and perhaps severely limiting his market (!):


"On intention, in short, is to make the Irish Times a first rate Irish newspaper, complete in its details, sagacious and consistent in its policy, and faithfully reflecting the opinions of the most independent, intelligent, and truly progressive portion of Irish society."


Social newsworking has the intriguing possibility to combine focussed audiences with a wider democratic reflection on the presentation and analysis of news, than can today's newspapers which are hampered by advertising based and subscription business models and in almost all cases, their shareholder returns to their owners.