Is Powerset capable of dethroning Google ? Only time can tell. For now, Powerset is facing problem keeping its promise and now undergoes management change to appease its investor. Not surprisingly, Techcrunch has been monitoring its progress since the company formation, and among the blog written is PowerHype At Powerset where Mike quote PowerSet as
in some ways they are very secretive – everyone who gets to see “The Demo” (as it’s now being called) have to sign a nondisclosure agreement.
People who’ve seen The Demo say it is a completely controlled environment. The index is limited to just a very small sample of high quality sites like the New York Times website, and the search queries are driven by Powerset employees as well. With that kind of setup, almost anyone could show a stunning demo.
Tracking its progress, VentureBeat post an article today, Powerset, the hyped search engine company, sees shakeup, that reveals the problems plaguing the ‘Next Big Thing’.
Powerset, the “natural language” search engine company saying it wants to take on Google, has seen a shake up in its management.
Somehow, they attribute one main delay is due to licensing agreement with PARC. However, it really doubtful if PARC technology could really be the solution and possibly be the worst problem because it yet to prove itself in anyway, let alone handle the real world with aplomb. If PARC technology is really so good, why isn’t it develop this further themselves through a spin-off ? Hasn’t PARC learn the lesson from the days of Apple where PARC failed to commericalise the GUI idea only to let Apple capitalize and profit on their GUI invention ?
I give Powerset a benefit of doubt that this is the reason and given that they have hired many intelligent ex-yahoo’er (see PowerHype At Powerset), they just might pulled it off, but remember that they are not developing simplistic web portal technology but extremely complex search engine ever built, one that is naturally intelligent enough to knock Google off.
“This is really hard, what we’re doing,” Pell said, of the company’s ambitious goals. He said one cause of the delay was the time it took for Powerset to license key natural language technology from PARC. Powerset finally got access to PARC’s source code in January this year.
The worse is yet to come, and PowerSet has to be answerable to its investor over the large sum of investment. It doesn’t sound too good when they try to raise more money despite not having anything shown to the public. They may want to maintain secrecy like Google but it may backfire since they have no traffic or monetary result to speak of, only controlled search result. Will throwing money solve the problem ? But what problem are there ? I think the problem lies in their ambition, and no money can ever solve it.
Powerset has also been trying raise money from venture capitalists at a very high valuation, and the shakeup suggests it wasn’t getting very far in its efforts.
This is instructive for founders and entrepreneurs who try to raise money early at stratospheric levels: Powerset last year raised $12.5 million last year from Foundation Capital, the Founders Fund and long list of individuals, giving it a post-money valuation of $42.5 million. Now, in order to raise money again, the company needs to seek a higher valuation, so that Foundation Capital and the other investors feel they are getting their money’s worth. However, that’s hard to do unless you’ve shown you can perform on your plan. We’ve mentioned Powerset’s various product releases, but none show that it is close to prime time — a year after the investment by Foundation et. al.
It look like Powerset is scaling down its ambition for the time being crawling the Wikipedia instead. At least it is better than nothing. But still with Wikipedia growing fast each day, when Powerset is going to see the light of the day ?
The company is still not crawling and indexing the entire Web for its search engine. It is still focused mainly on an index Wikipedia, a site that has clear structure and relationships between objects and their definitions. This gives Powerset a nice testing ground for its product. But the wider web is much more complex, and Powerset has yet to tackle it. In part, Powerset has been hampered by limited resources, Pell said.
Despite having some doubt about PowerSet’s promise, it still too early to tell if they can still continue the momentum. Powerset’s fate is very important to the progress of natural search development. If Powerset’s life is sealed, it will definitely impact the future and investment of similar search engine.
PowerSet’s ambition to take on Google could probably be forced by circumstances rather than by true intention. Money is unlikely to be thrown at being the second best but to be the number one. Still whatever intention PowerSet has, it only do the world a great service by keeping search innovation alive.
Google versus Mahalo versus Ask.com
To test how well Google fare against the best search engines, I use a human-assisted search engine which purport to give meaningful result and another based on hybrid natural search engine. I use the search phrase why study for an mba for testing and the result speak for itself.
Then I use against , the next best search engine that incorporates natural language technology. I find that Google provide the answer with most of the keywords whereas Ask.com tend to produce unrelated answer like GMAT, course etc. Perhaps Ask.com is too intelligent for its own good.
Why Google remains superior to natural language ? There are several reasons why Google will continue to reign superiority over natural language search (NLS) engine.
First, Google maintains a updated indexes of every words in a webpage without the need to understand the contextual meaning behind the words, and thus doesn’t need to establish association with any word. This make it able to search virtually any language without the need to understand anything about the language since it just indexes the word as it is ! Performance is faster because less processing is needed albeit at expense of disk space. Natural language search (NLS) depend more on efficient algorithm to reduce space requirement and perform contextual search however it require great deal of processing power. Moreover, the latter will have to be tuned for different language like Chinese, Japanese, which only increase complexity and processing time. Yet, NLS’s space requirement could even be worse if it use caching to improve performance since a lot of association/relationship could be made between word. Nevertheless, NLS will be very useful in controlled environment where search is optimized for specific purpose rather than general purpose like Google.
Secondly, contextual search might even be a solution in search of a problem if apply to similar scale like Google. Consider that most sites tend to use simple and commonly used phase that most user can relate and familiar with, Google may easily return these results quickly. On the other hand, NLS might need to analyse the meaning behind the search phase and then return the result. Even if it is able to cache commonly used phase, wouldn’t it’s contextual capability been under-utilized then ? Are NLS trying to solve 10% of the problem where Google is solving 90% of the problem ? So how can NLS ever outperform traditional search engine like Google ? It will be better to leave Google alone and fixing thing that isn’t broken or not worth fixing. NLS is better off concentrate on concrete achievable goals before it will want to take on the giant. The industry will do NLS a great disservice if NLS is perceived as a hype due to millions of investment money throw at the development without a concrete result. Indeed, it is not the failure of NLS, but a failure to align ambition with reality. A dream can only be achieved with concrete small step not with one single big swoop !
Another reason is what if use enter some keywords that has no meaning contextually, for example, mba why for study. NLS might probably waste time processing garbage and return garbage as a result. Not many user will enter phase for searching, and most will use keyword instead.
Although, a hybrid solution which combine the best of traditional search with Natural Search like Ask.com seem to give the best compromise in theory, but by judging the search result of Ask.com, it definitely has a very long way to go. Currently it does not have any distinctive advantage to speak of.
Nonetheless, NLS is still a important area for research as it’s potential grow with increasing power of processor and hardware.
Concluding, I still will like to wish PowerSet all the best in its endeavour in developing the most intelligent search engine ever. Only by trying the impossible can we see breakthrough.