A Rant on Arc Search

The author says “The solution to search is a search engine that is aligned with the users.”

I wholeheartedly agree.

The incentives at play are often at the detriment of the person searching, and for the benefit of the company and shareholders of the search engine. Google measures ad revenue and related metrics far more than the abstract concept of whether people are getting the best search results.

The Browser Company, with 22 investors putting in over $50M, has pressure to make a return on that money. The way to do that in the long run? Show paid results. Unless they can make a subscription model work, which many search engine startups have failed at, they'll eventually resort to paid results. Then people will see what's been paid to show up, even if there might be something more relevant to their search than the ad.

What would a good search look like?

This isn't the first time I've lived through a "search is broken!" era. Search has stopped working for me at least 3 times before: 1. When the primitive methods we had for searching usenet were overwhelmed by the growth of the internet. 2. When Yahoo finally had to admit that human curation was not going to be able to keep up with the growth of the web. 3. When Alta Vista just could not fit an index to the web inside of one computer, no matter how powerful.

This current era we are in feels a lot like those old transitional eras did. In each case, the solution was to invent some kind of gee-whiz technology to index and search ever larger data sets.

We now have the gee-whiz technology, but what's different this time is that it is being used to create yet more content, on every conceivable subject, and with every conceivable degree of connection (or disconnection) to the ground truth.

Its really hard to see how the problem gets solved this time, but then again, it was hard to see how it was going to get solved all the previous times.