“Trust is not a calculation”

I agree with Jeff Jarvis on this point. I wrote the following in a comment to Jeff’s post:

The “don’t be evil” slogan is either naive or disingenuous on Google’s part, because it’s based on a presumption that concentrated power can be managed by good intentions. Historically this proposition has rarely if ever been true. What really keeps power in check is accountability based on competition and openness. Google has competitors but the essence of Google’s position on openness is, “trust us.”

This is also the problem with Google’s TrustRank scheme. Rather than merely evaluate the trustworthiness of news providers on a case-by-case basis, as we do now, under TrustRank customers would also have to evaluate Google’s judgment in deciding which news sources are trustworthy. This might make sense for customers who share Google management’s biases but for everyone else it’s a burden.

Google is a network company par excellence. It knows how to add value by maximizing network opportunities for content providers. Where Google stumbles is in trying to add value by providing or managing content according to its own values hierarchy.

From what I can infer, Google’s TrustRank system is mainly a branding scheme — in this case, the brand is based on an automated method for vetting content according to Google management’s preferences. It sounds like NPR without the overhead. There’s nothing wrong with that but it isn’t special either. All news aggregators impose their biases to some extent, because they have to decide what’s news and what isn’t. Google’s proposal, if it is what it appears to be, amounts to another aggregation scheme, but one overlain with hubris about what used to be called “scientific methods” — i.e., the idea that by mechanizing selection rules one can remove human bias. It will be interesting to see what emerges — if Google actually implements the TrustRank concept — and how it fares in the marketplace.

UPDATE: Read Tim Oren’s comment on Jeff’s post for another view.

UPDATE2: Tim Oren’s second comment adds more reasons to be skeptical about the validity of my concerns.