The New Yorker published a cartoon in 1993 which showed one dog sitting at a computer terminal saying to another dog, “On the Internet, nobody knows you’re a dog.” A few years ago, I attended a presentation given by one of the computer graphics experts who worked on the movie Jurassic Park. He described how they created many of the dinosaurs completely by computer. He continued by saying that within a decade, they will be able to create human beings on film completely by computer without any need for actors. The power of Photoshop to recreate photographic reality is frequently demonstrated as magazines are caught digitally manipulating images to meet their needs.
Over 15 years after the original New Yorker cartoon, there is still no widely deployed mechanism for verifying the identity of anyone or authenticity of anything found on the Web. As the Web becomes the primary source of information for more and more of the world’s populace, it becomes increasingly difficult to discern truth from fiction. The question is, “Whom do you trust?”
Over time, as technology has evolved, new trust models have been developed. Recommendations through friends – This is perhaps the oldest method of establishing trust. You simply ask someone you trust for a recommendation, e.g. you move to a new city and ask a colleague to recommend a doctor or attorney. You believe the story because you know and trust the source.
- Recommendations through trusted third parties – Restaurant, movie, or wine reviews in a newspaper are examples of this model. Because you trust the judgment of the reviewer, you trust their recommendations. Gartner evaluates IT products and vendors, and their ratings of consulting firms demonstrate the effectiveness of this method.
- Process creates trust – Traditional journalism requires the validation of a story from more than one source. You believe what you read in the New York Times because you trust the vetting process they use before they print a story. Wikipedia is also an example of this trust model. You trust the contents of Wikipedia because you believe that the “crowd sourcing” process is effective.
- Community ratings – Zagat guides demonstrate the effectiveness of this approach. Rather than depending on a single trusted third party, you simply aggregate the opinions of a large number of people to create a recommendation. Based on their success with restaurants, Zagat extended their model to hotels, nightlife, movies, music and now even dating (and dumping). This model has been dramatically expanded on the Web to everything from local repair shops to attorneys and doctors.
- Reputation systems – eBay’s trust model is perhaps the most novel. With most eBay transactions, an auction winner sends payment to a completely unknown seller. The seller then ships the product to the winner. There is no formal recourse if the product does not meet the buyer’s expectations or even if the seller never ships the product at all. Within eBay, the system of community reputation encourages buyers to rate sellers. For a prospective buyer, a high reputation score equates to a quality seller with satisfied customers, which means they can be trusted.
However, as information continues to explode and search engines now include results from Twitter and Facebook, clearly a new trust model is needed. Recently David Pogue, the respected New York Times columnist, was accused of a conflict of interest by several Twitter posters. One such Twitter post was from a user with the name “John C. Dvorak,” which also happens to be the name of another well respected computer journalist. David Pogue gave an interview about the incident and took John Dvorak to task for his Twitter posts. Unfortunately, the Twitter poster was not the computer journalist John C. Dvorak but someone else with the same name. The journalist actually posts under the Twitter name “TheRealDvorak” and had made no comment at all about Pogue. In this case even Pogue, an experienced New York Times Reporter, didn’t realize he had mistakenly assumed the identity of the poster.
Twitter responded to the growing problem of mistaken identity by providing a program which attempts to verify the identity of some Twitter users. Unfortunately, the program is limited to a very small number of celebrities and given the rate at which Twitter is growing and their limited resources, may not be expanded any time soon. However this problem will continue to grow as more and more people believe what they read on Twitter.
Solving this problem represents a great challenge which will require significant new innovations. If you can’t tell who’s who, or what’s what on the internet, its value as an information repository will begin to diminish.
For more of Sheldon’s thoughts on these topics, check out his video interview on IdeasProject.