I wrote a comment piece for this month’s Ideas Bank in the UK edition of Wired about trust, reputation and the computer algorithms that try to recreate them in a meaningful way.

Perhaps you can tell by the title, I don’t think any of them are there yet.

Here’s a snippet:

…according to the research, our past actions can’t predict future intentions. They can only signal our past reliability and suggest what we might do in the future. Regardless, an entire economy of ideas is based on online reputation, because it’s the people with the reputations and the social currency who get people’s attention. “Attention” has become a synonym for “influence”. And that’s nonsense.

Furthermore, I don’t believe in “good enough” solutions that try to reduce the complexity and messiness of humans into 1s and 0s. That’s what the serendipity engine project is trying to get at, and what my research over the next year and a half is seeking to unpick.

I also really dislike sentiment analysis.

Read on..