A lot of the excitement around Qwiki is because of its ability to generate media on the fly that combines text, audio, and animated photos.
Right. Qwiki recently announced closing $8M in financing, and this necessitated a rethink on my part. Someone had given them $8M dollars, so perhaps there is more there now than I saw last time I looked at it, sometime in October.
And… nope. I have no idea how this company raised $8M, and could possibly be deserving of it in any universe, other than Silicon Valley. There is simply nothing there of substance.
I entered three terms that I know a little bit about: 1) Vancouver; 2) friendship; and 3) the French Revolution. These are three distinct object types: a place, an abstract concept, and an event, in this case historical. These are softballs for a service like this, supposedly in development for over a year now. I choose them for this post before I entered them to see what they displayed.
For Vancouver, Qwiki read aloud the first 1.5 paragraphs of Vancouver’s Wikipedia entry, and then a few selections from the second paragraph, and then a completely irrelevant section, for an introduction to Vancouver, about the Vancouver School of Conceptual Photography, also on the Wikipedia entry, near the bottom. Most of the photos displayed were taken from Fotopedia.
For friendship, Qwiki read aloud the first paragraph of the friendship entry on Wikipedia, a simple definition in contrast to other relationship types, and then a bizarre inclusion of a reference to an obscure and irrelevant academic study. The pictures were of very cute little dogs, and Purdue University, both taken from Wikipedia.
For the French Revolution, Qwiki read aloud the first paragraph, the first sentence of the second paragraph, also from the Wikipedia entry, word for word. The images displayed were again from the Wikipedia entry.
In all three cases the related links were generated from Wikipedia “See Also” links. This is why on something like friendship you get a link to autism in animals, not something anyone will likely care about, if that is even a real condition. Qwiki did better on suggested items for Vancouver and French Revolution given the richness of the material, but still, nothing more than reprinted Wikipedia See Alsos.
The implementation of something like this is not technically overwhelming, or technically defensible, as demonstrated by fqwiki. The image processing is trivial, if tedious, and all the data is already organized and available to anybody in a structured format. They are only presenting the Wikipedia database.
1. How is this worth a valuation able to absorb $8M and not dilute the founders beyond motivation? It is not, obviously.
2. Why do you need $8M dollars for an alpha consumer product which has no real technology of its own to create? You don’t, unless the only barrier you can possibly erect is marketing.
3. What is to stop Wikipedia, the actual curators of the real value, from adding the exact same audio and picture show to every entry? Why then would I visit Qwiki ever again? Absolutely nothing, and few would.
The Qwiki product is also not very useful. As many have pointed out, it is a very inefficient way to learn information. There is no text to read while listening. After you consume more than a few entries the tacky image animations actually become very irritating; they actually started to give me a headache (it would be much better to do a cross fade on a consistent image size). The trend of the Qwiki Compete graph is not a good sign, either. Qwiki has virtually no retention.
Qwiki is little more than a toy, basically, as any cursory investigation reveals. But this is Silicon Valley, after all. Let’s give them $8M dollars anyway, just in case.