Three weeks ago, Tucson made headlines because the Arizona Wildcats came out of a weak season to capture the NCAA (College Basketball) Championship. Cab drivers are often leading indicators - three weeks ago, mine was glued to the quarter-final game as he drove me to the posh desert resort in Tucson AZ, where Esther Dyson was about to host her 20th annual PC Forum.
PC Forum is Big League, unlike the NCAA. On the first night, for example, I ran across FCC Chairman Reed Hundt, with his wife and children. The, "Hi, David," "Hi Reed," encounter (the Chairman did sneak a peek at my name tag) wouldn't have been possible in Washington, DC's corridors of power. At PC Forum, not only Hundt, but new AT&T President John Walter was to speak, as were Michael Bloomberg (Bloomberg financial information machines, Bloomberg Information Radio 1130AM), John Reed, (CEO, Citibank), and a variety of equally heavy, though somewhat lesser known, thinkers.
The "official" theme of the meeting was "Models and Metaphors." George Lakoff, the UC Berkeley linguist, was the poster-professor for this theme. He spoke on the use of metaphor in everyday life, and of its power to direct (if not control) thinking. For example, saying that a relationship is at a "crossroads" implies different actions than saying that it is at a "dead end," though in both cases the relationship is seen as a vehicle or a journey. By extension, thinking of Internet content as occurring on a "page" implies different actions than if it occurs on a "site" or a "channel," or if it is seen as a media-less "event."
Metaphor is ubiquitous in information technology (e.g., file, desktop, storage, browsing, push, window, . . .). Metaphors that Lakoff introduced for service businesses included:
New insight can come from deconstructing these and examining the underlying concepts. Which of the above remind you of AT&T's business models? Which don't?
|
"Ad agencies are NOT on the side of finding
out that nobody's watching." |
Michael Bloomberg was very down on traditional advertising. He said that the "problem" with the Web (for traditional advertisers) was that you CAN measure the response. The advantage of the other (mass) media is that nobody knows.
Lest we think that "open systems" is some kind of universal principle, Bloomberg has made quite a business out of selling his proprietary, closed system - 73,000 users at $1200 per month, growing at 30%/year. As sources of information develop on the Internet and as Bloomberg migrates certain services to the net, it will be interesting to see whether the proprietary services evolve apace.
Esther Dyson introduced Reed Hundt's talk on providing Internet in every school. Before Hundt spoke, Esther asked the audience how many people thought there were big problems with education - virtually the whole audience raised their hands. Then she asked how many people thought the Government should fix it - only two hands went up. Then she asked how many people would like to do it themselves, and there was a smattering of hands. So how's it going to get done - or is it???
|
"Listen, parents are still a really good idea." |
Then Hundt began with the Parable of the Hot Dogs and Buns: there is a hot dog for every bun, with no Hot Dog Commission, no bun subsidy, etc. It was refreshing to witness a regulator being self conscious about the role of regulation. Hundt claimed that by last count there were 2850 schools already on the web. He described an FCC proposal, to be heard on May 8, whereby a $2.25B fund would be created by universal service fees from large telcos, to be supplemented by $2.25B from local education budgets.
I do not know AT&T's position on this, but under an Internet World scenario, my personal opinion is that it would be good public relations, if not good business, for AT&T to lend public, visible support.
Robert "Internet Meltdown" Metcalfe spoke against the proposed FCC program. He was concerned that if ISPs took this federal money to wire up schools, that the next step would be federal regulation of ISPs. He said, "In business, if something doesn't work, you stop it. In Government, if something doesn't work, you redouble your efforts." Later, somebody who knew Metcalfe remarked to me, "Bob never met a federal program that he liked."
Peter Schwartz, from the audience, commented that one value of the Internet was to re-connect teachers, kids, and parents to schools and to each other - that the most powerful impetus to learning comes from home, from parents, and that Internet connectivity would be a way to get parents back into the loop with teachers and kids again.
Somebody (Peter Schwartz?) suggested that Internet Access could become an employee benefit. It certainly is moving in that direction (recent Lucent and AT&T dicta notwithstanding) and some tax breaks would help it along. (And why not? Talk about non-coincident peak usage . . . ) Such a move could be seen as a strong leading indicator for an Internet World scenario.
In the next panel, John Reed (CEO, Citibank) spoke eloquently about trust and risk. The biggest risk that Citibank faces is "Reputation Risk" - the public perception that the bank is "not secure," or, even worse, "not trustworthy." In other lesser developed countries, the words "American Bank" have a very powerful connotation of longevity, financial stability, low inflation. Citibank spends much energy allaying government requests for more disclosure, and more "cooperation, and Reed made it clear that Citibank will comply with court orders, but that is where it draws the line - if it loses the trust of its customers, it goes out of business.
Also, somewhere in PC Forum, five elements of Trusted Systems - Human, Political, Machine, or otherwise - were enumerated:
|
Hey, this is important! |
To a first approximation, this set seems right to me. Maybe you need some others (e.g., fairness, consistency). Recourse was a solid insight for me, though - if you dial a wrong number, or have a "funny" item on your bill, you can call AT&T for an immediate adjustment. This reinforces customer trust. Even if you never do it, knowing that you can is important. Same with credit cards, where the mediation of disputes - the so-called Privatization of Small Claims Court - is key. Maybe the Government is a non-trusted system these days because its recourse systems are broken. But whatever else you might think, the warm and fuzzy feeling behind a brand is less about image and more about direct experience in the Information Age.
One discussion of community drew Stewart Brand to the floor mike. He said that the Internet was causing a change in the power relationship between corporation and customer - "At the complaint department, people used to be all alone." He also said that communities are, by their nature, conservative. My thinking about this extrapolates like this: to the extent that communities are strong, they include the in-crowd and exclude the out-crowd, they have shared norms and values, and certain economies of thinking (group-think) that are slow to form and slow to change because they require consensus. The iso-transaction-cost (or iso-pain) contour is the path of least resistance.
During one panel, Steve Case (CEO, AOL) described an innovative relationship between telco "Telsave" and AOL - where immediate on-line call detail was available via AOL, and billing was to your credit card. Gentle reader, if you are an AOL subscriber and you know more about this, or can find out, please let me know!
There were several demo areas - pods in the PC Forum vernacular - where different families of applications were being demonstrated. In the Social Computing pod they had discovered PHOAKS (by Loren Terveen & Will Hill) all by themselves. Ron Brachman was very proud (so was I). But why are we hiding such lights under the bushels? (Maybe somebody should have known that this would be considered cool by this audience, and been more proactive about it.)
One demo represents an important early example of the kind of high-value voice-enabled Internet application that, when they begin to appear in numbers, are likely to make POTS the technology that you use only when Internet is not available. Placeware (www.placeware.com) - a Xerox PARC spin-out - has conferencing software that allows multi way voice interaction among a speaker and members of a multi-location, multi-party audience, plus graphical presentation, plus some (fairly impoverished) graphical audience feedback and identification. The total experience is less than "being there" - nothing exceptional or gee-whiz - but it is much more than video-conferencing or audio conferencing, and easier to improve and extend, too!
Throughout, in her introductions and remarks, Esther used the metaphor of a garden - with images of deliberate planting, diligent weeding, regular watering, abundant harvest - to describe the Net, and indeed, PC Forum itself. Now head gardener Esther Dyson herself addressed the audience on Labels and Disclosure. She talked about PICS, the voluntary content labeling protocol (she acknowledged Paul Resnick of AT&T Labs for leading the effort) and related concepts. PICS is good because it is (a) voluntary and (b) richly flexible, therefore much better than censorship. She said that people wanted schools to be more like markets because markets have feedback loops and are self-correcting. George Lakoff, in an invited comment, remarked that the business metaphor for education was "a disaster" because if a business makes mistakes and fails, or a product fails, that is good in terms of the overall market, but schools MUST NOT fail - therefore they need to have other, much more specific, self-correction mechanisms. Furthermore, he says, education is not a product, and teachers are not "education delivery systems." Metaphors really do shape how we think.
Christine Varney - a Federal Trade Commissioner - was, like Hundt, self-conscious about her role as federal regulator. The FTC is responsible for federal level issues like "truth-in-advertising," fraud, and competitive practices. She said that if there are existing laws on the books, use them; if there is a way to get redress in the courts, use it, but if not, then let's talk about the necessity of new regulatory approaches. She spends a lot of time explaining to the non-networked that most so-called Internet crimes are really just crimes. On privacy, she advocated voluntary disclosure (e.g., in the PICS framework) of how commercial concerns on the network would use personal information - if you know that e.g., LL Bean will sell your name to others, but e.g., Eddie Bauer will not, maybe this fact will become a competitive distinction. Varney was impressive for her sharpness, her articulate manner, and her humanity.
There was a mercifully brief session on infrastructure, starring Tom Jermoluk, the president of @Home - the cable-data company - who seemed uncomfortable and defensive about @Home's sliding goals. He reported with some mix of pride and relief on a late 1996 project that has put high speed Internet connections into 94 units of a 98 unit apartment building in LA - using standard vanilla ethernet. He defended the need for "telco return" (i.e., the current inability of cable plants to really support two-way) by saying that @Home was "a 4-5 year project" and that they anticipated making progress on two-way cable shortly. Telco return "is a stopgap," he said.
Somebody in this panel said that of the 5000 homes in Los Altos Hills (posh community in Silicon Valley) that 1000 have at least one T-1 going to it.
|
"No CEO enjoys saying 'Trust me on this one,'
to his Board of Directors." |
Then John Walter, the unknown new President of a small, fast, smart communications company, spoke. He had just come in from Madrid, London, and Tokyo, but he did not say what he had been doing there. He said that AT&T was not a phone company, that AT&T was not traditional, that this was not corp-speak, and that we were not in deep denial. Rather these were the new business realities. He said that we must jettison old engineering and business models and embrace models that anticipate change. He said that AT&T would become a Communications company, not a Telecommunications Company.
Walter said that the new AT&T would be an "ecosystem of partnerships, relationships, alliances . . . a hybrid of industries . . . creative coalitions." He said that the recently announced AT&T fixed wireless technology was technologically proven, but that there were still economic issues. But, he said, fixed wireless was "just one way to break the paradigm" of the RBOC control of the last mile.
In another reference to the value of technology, he mentioned that technology was important in AT&T's success, and said, gesturing, "That is why I have Dave Nagel here today."
On Corporate Culture, he said that he would "shorten the line of communication." He said that the Business Unit idea had run its course, that AT&T was going to be a "Market Facing Organization," with centralized control of the brand. He emphasized the role of "management issues."
Walter's prepared remarks lasted maybe only 10 minutes - then there was the Q&A, started by Jerry Michaelski, who said how much he loves AT&T, how he owns AT&T 500 service, 800 service, wireless service, how he is PICed to AT&T at home, has a Universal Card, and asked what were John Walter's plans to integrate all of these many elements. John's reply again touched on management issues - he explicitly invoked the name of Lou Gerstner. He spoke of the need for aligning AT&T's capabilities more effectively. He continued, "But we cannot build and control everything," and repeated that AT&T needs partners that add value to the proposition.
I particularly liked one crisp exchange:
Q: "When will I be able to get a T-1 to my house?"
A: "When there is REAL local competition."
The final talk of PC Forum was by Tim Berners-Lee, the inventor of the Web. He talked about the use of labels - metadata (data-about-data). He said that computers browsing the web, "was part of the dream" that he originally had. The dream, that the Web would be a place for every kind of information, from personal information, medical info, datebooks, to family information (photo albums, etc., team, school group information, business, town, and public information, all labeled with verifiable authorship and content labels, appropriately protected from those who should not have access, and accessible to those who should. I sensed his great pride at how far the world had come, and his great awe of the vast space yet to cover.
Berners-Lee got me thinking about PICS again. Most of the cost of the PICS system accrues where human judgment must be applied - in the labeling of the content. True, for some of the more trivial kinds of content labeling algorithms can be used, but one joke at PC Forum went, "Remember Artificial Intelligence? - That's where you have an IF statement." So non-trivial third-party distinctions, say whether content is about obstetrics or pornography (or pornography disguised as obstetrics), or whether content is "suitable for 14 year olds," or whether to classify content as "entertaining" or "instructive," or "Christian" is likely to require human judgment. And, I dare say that one would not want to leave such judgment to low-wage batch laborers - this would be highly skilled labor. All but the most trivial labels are likely to be expensive.
Therefore, given the significant cost of rating a web site, I fear that PICS could start a vicious cycle leading to the re-massification of media. Here is how this could work: (a) third-party raters will tend not to rate sites that they consider marginal or that do not get a lot of hits, (b) people whose web access is guided (or determined) by a third-party rating service would not see unrated sites, (c) unrated sites would get even fewer hits, (d) see (a) above. A nontrivial number of Internet users are apt to have their third party rating service chosen by their employer, their ISP, or their government. Presto, the re-massification of media. Or as one of my colleagues remarked, "Before you know it, you are back to three networks again."
|
"Nobody does betas anymore." |
In truth, it probably would not be that bad. And voluntary content filtering is definitely better than censorship. But content labeling and/or filtering is virtually certain to become a very important form of Internet business - perhaps it will be to Internet content what operating systems were to PCs. There is huge demand, even though none of the value is kinetic yet. And the purveyance of such editorial metadata constitutes a huge opportunity for AT&T, by virtue of its trusted brand and respected reputation. But as a future customer, and perhaps a victim, I hope that the Internet community will be able to think its way towards preserving the diversity and openness that have made it great to date.
There were other aspects of the session that made PC Forum a great experience.
Three weeks later my brain is still buzzing - it must have been an OK meeting!
David S. Isenberg
Date last modified: 23 Dec 1997