Instant search results, presented as I type? This is what amounts to a reinvention of search? I'm not buying it, and neither should you.

Yesterday, Google (Nasdaq: GOOG) unveiled Google Instant at a press event at the San Francisco Museum of Modern Art. Press who were at the event say executives described the new approach as psychic, as if Google knows what you want before you do.

I've got three reactions to this.

First, ewwwwww. Creepy.

Second, while I understand the importance of speed, what I really want from Google is better results. I want it to see the patterns I'm seeing, and fetch data accordingly.

For example, if I sense that cloud computing is catching on in Asia and enter "cloud computing Asia" into Google, I want results organized by statistics, corporate happenings, research reports, and the like. Google should allow me to sharpen results and get closer to the answers I need. Today, that's work I do on my own.

In this sense, Google made a mistake by making Instant big news. Rather than excite users and investors, it has made clear just how far we have to go in making search better -- and in the process created an opportunity for Microsoft (Nasdaq: MSFT), AOL (NYSE: AOL), Ask.com, and foreign rivals such as Baidu (Nasdaq: BIDU).

Third and finally, Google Instant's great strength is discoverability. Watching Google work in real time allows me to see not just a shifting mix of results, but also a wide variety of queries, some of which were better than the queries I've come up with on my own. And better questions should lead to better results.

Google Instant is neither a breakthrough nor a failure. It's a small step forward on a very long journey to make search more precise than it is today.

Now it's your turn to weigh in. Did Google Instant meet, exceed, or miss your expectations? Join the conversation in the comments box below.