Log In | Register
So apparently there's this new search engine out there called Cuil.
The brainchild of Anna Patterson, an ex-Google engineer, Cuil is an effort by four people (two of the others also worked at Google) to make a better search engine. Not only does it protect your privacy, it also searches through 120 billion web pages (which may or may not be more than Google indexes, nobody knows), and presents the information in a "magazine-like" format with images. Also, they claim to have more relevant search results based on page content rather than rankings. I tried Cuil out, and found out that the three claims mentioned two sentences ago are most definitely true. Those three at least.
Searching the vine for references to cite here, I found only one other article about user experience with Cuil written by Steve1151 here. Steve, my friend, I commend you, but do you really have to be that soft on them?
How bad is Cuil? Let me tell you.
I first went to Cuil.com after reading an article about it. A simple, clean page with a search box. Looked nice. I thought of what to type in.
A bit of background on me: as some may guess by my purple fuzzy fox avatar, I'm a member of a subculture called "furry." It has to do with animals being people and stuff, I won't get into it, but anyway, that's the first thing that came to mind. I typed the word "furry" without quotes.
I would have been perfectly happy had Cuil showed me some stuff on cute and cuddly furry animals and nothing related to my somewhat niche interest. After all, that's what most people think of when they hear the word "furry." A second or two after clicking the search button, I was absolutely astonished at what I saw.
We didn't find any results for "furry"
Some reasons might be...
* a typo. Please check your spelling.
* your search includes a term that is very rare. Try to find a more common substitute.
* too many search terms. Please try fewer terms.Finally, try to think of different words to describe your search.
Wow. What?
Last I checked, "furry" was an English adjective meaning "covered with fur." You'd think maybe this word would appear somewhere on the Internet in some fashion, or at least something related to it. Not according to Cuil.
I got the same result for "fur." This is not turning out well.
To be fair, I searched for something a bit more common, "wikipedia." Ahh, now we're getting somewhere. I finally get to look at Cuil's "magazine format." It looks quite good. Little snippets of text with images next to them, as well as some related searches I might want to try, glommed together in a nice, Web 2.0y fashion. Let's see what the results are, though.
"The word on Wikipedia: Trust but verify"
"Online Wikipedia is not Brittanica, but it's close..."
"Critical views of Wikipedia - WIkinfo"
Okay, where is this Wikipedia thing everyone's talking about? Ah, here we go:
"Jennifer Lopez - Wikipedia, the free encyclopedia"
Really? J Lo? That's the most important Wikipedia page according to how things are listed on Cuil search? Whatever happened to, say, THE HOME PAGE? And why is it that there are "78,600,000 results for wikipedia" but you only have four pages of results, some of which list the same things as the previous pages?
Just as a final test, I typed "cuil" into the search box. I got a seemingly endless amount of pages talking about towns in Scotland. Although for some reason, more pages appeared when I clicked on the next page, which I figured was supposed to happen. This system seems more inclined to display the supposed 121,578 results for "cuil," but it didn't work when I searched for wikipedia. But let's not get distracted by that. Can we focus on the fact that this search engine doesn't know about its own existence?
Obviously, they're suffering from a little first-day syndrome. Searches were occasionally slow, and their About page produces a 404 error. But whenever a search does work, it, well, doesn't work. Or certainly not the way any normal Internet user would want it to. I don't know if there's some kind of arcane, mystical language it's looking for that would make it work (like, say, SQL). but the fact of the matter is that the average Internet user who likes Google is not going to find any reason to switch to Cuil. Hell, nobody would be interested in switching to Cuil if it runs like this.
This goes beyond a "launch problem," this is a design flaw of back-to-the-drawing-board caliber.
Or maybe it isn't. Maybe the algorithm depends on people doing searches and "training" the search engine to not suck. Well, it would be nice if you told us that, and maybe thrown in a "Pre-Alpha v0.00000000001" disclaimer for good measure. But the fact of the matter is that this search engine has been launched with every assumption that it's going to work, and it doesn't. This is absolutely unacceptable as far as Internet software goes. The only possible explanation I have for it is that it's high-brow New Yorker-esque satire commenting on everything we hate about the Internet.
UPDATE: A recent CNet article tells the tale about why things are going wrong: apparently Cuil's architecture is weird. When it gets more traffic than it can handle, it doesn't crash. It just sucks. Apparently it has a bunch of machines that index different categories, and when a particular category crashes, searches in that category start to yield useless results. It's a very interesting system, but frankly I'd prefer it work entirely or not at all.
Still, it does provide some closure as to why it's behaving this way: because it's seeing all these new, unknown people from all over the Internet and FREAKING OUT!!!! That said, a new website should never underestimate the power of news-hungry bloggers. Somebody needs to come up with a system of avoiding crashes, or in Cuil's case, nervous breakdowns.
