gettheanswer

Recently, we’ve launched a new function on  beta-version of our website. As we decided to do it quick and dirty, we have appreciated some feedback from beta-testers, if this function was sexy for users.

After a couple of days, we’ve realized that there was a small, but important flaw in introducing this element. The perception of purpose. As our website is stricly connected with q&a, this function was considered to be a part of the core. But it wasn’t. The aim of this, was to broaden the reason of the existence and sexiness of the website for users.

If we developed a final version and act with a big  marketing campaign after that, we wouldn’t have had enough time to improve a communication about this function, so it’d had been useless. Thanks to our quick and dirty lanuching on private beta, we are warned the we really need to describe it better. So we do  it, and  the usefulness is  soaring…

I’ve recently read a few articles at RWW about some aggregates of social feeds. These apps certainly can be useful when using multiple social networks and being interested in your friends activity - of course in case you have other connections in different networks.

Of course it can be not only joyful, but also useful for some people in some cases. For me it rather doesn’t work in this way. I’m not a fan of Twitter and other chatting. I’m rather interested in useful tools, helping with knowledge management, for example.

However, these services could be developed into such apps which would be able to help with daily life. For istance, they could gather not only twitter-like short messages ‘what are you doing right now?’ but also some useful data found by our friends during net-surfing. Of course, they’d have to allow the system to share that knowledge, but after they agreed to that, it’d be extremely useful.

I’ll give the example. If you’re searching the net and you’ve found something, you can put it into your twitter-like system and share with friends these links connected with your short description.

I hope it’d work

This document is prepared for my business English studies at the Warsaw School of Economics. We are supposed to present a debate on a particular topic, so that I chose a topic connected with a semantic web, or you can call it a future of the Internet. As it is a kind of debate, there are some thesis or views presented below. They are subjected to be discussed, so feel free to comment on them and present your views.

#1 The future of the Internet is Semantic Web

Today’s net is a messy set of data on separate webpages. The content is rather poor quality, often duplicated, spammed or simply wrong and unverified. The links between websites are often irrelevant, not up-to-date and used too rarely. Thus, the published content is like a jigsaw-puzzle, it’s not useful until you do a jigsaw, what can tike a long time.
On the contrary, the semantic web will be a network of logically connected pieces of knowledge, easy to obtain the proper and verified content easily, using validated and frequent links between websites.

#2 There could be a better way to search than Google

You can theoretically find everything in the Internet using Google, because it has indexed almost all the published content and it uses great algorithms to present the proper information. On the other way, there are some obstacles: you need to type the relevant keywords; this search engine doesn’t understand neither your needs nor meanings of the keywords in the context. To provide you with the proper search results, any engine needs to understand your present needs, all important circumstances, a little bit of your personality and an objective of your search. Everything above consist in the semantic (usage of) web.

#3 There is a huge need from the world society to provide them with better results of finding knowledge

All funny (and often useful) websites, such as YouTube, Facebook, Myspace, Twitter and other deliver with a big portion of entertainment but they don’t solve the problem of having too little time, too many obstacles and too poor results of searching for the reliable knowledge for business or educational reasons. People want and need to receive answers to their questions and find needed knowledge both quickly and easily. It cannot be done by using today’s techniques simply because a lot of results in Google are and will be irrelevant and useless in particular cases.

So, are there any possible technical threats that semantic web will fail?

#4 There would be a bottom-up development of Semantic Web

As Wikipedia started to organize knowledge from creating articles one by one, the same Semantic Web will grow up thanks to knowledge organized piece by piece. In my opinion, it’s the only possible way to challenge the problem of huge store of untidied information. The trials of “packaging” information with so called “meta-languages” is extremely difficult, so I see the future in organizing valuable knowledge chosen by people, in particular set of data (a webpage) by eager users, for their purposes.

#5 It’s really feasible

A term ’semantic web’ seems to be like a dream or simply a vague definition. Difficulties with this term becoming reality are basically dependent on the fact that a lot of webpages’ owners have to cooperate together. This can be omitted, however, by creating bottom-up strategy of development semantic web, that is focusing on chosen aspects, analyzed, tested, and developed by selected users to create the beginnings of the semantic database.

#6 You can control your privacy

If nodoby does it, Google will do it. The best solution for each of us, there is a development of one of the latest start-up’s, which enables users to organize, share and collect the knowledge in a convenient way, using achievements of semantics, social graph (a structure of connections) and nice interfaces. Such new services, new webpages would be used only for these purposes and handling these kind of (partially private) information would be a responsibility of company different than Google. It’s less dangerous for us, because of the threat the Google would ‘know everything about us’.

#7 There can be really daily life usage

Eventually, you can use a couple of different services. Nothing is assigned to fulfill each and every need. So, the samples of semantic usage could be not necessarily to substitute Google, but be a supplement in such cases, where using traditional search engine is useless or time-consuming.

#8 It’s an evolution, not a revolution

It’s a constant change. One of the first step is Wikipedia, which organizes, shares and connects human knowledge. The second one is a ’social-graph’, social-networking websites, structure of our friendships, connects etc. Then we will develop tools for gathering the knowledge for later usage, for sharing with friends and other members of our communities and to use (with permission) from other ‘intelligent databases’.

As a non-native English speaker, I got used to using a word “sexy” as an adjective to anything hardly connected with intimacy and relationships… Actually, I like this kind of defining feeling about something what can be serious and comprehensive in business matters. It makes the artificial, scientific or technological staff more human, more understandable and daily.

In a traditional way, ’sexy’ meant an attractive person (esp. women), so this was about feeling to another human, attractiveness to someone - having big potential of useful some aspects of life Then, ’sexy’ was introduced as any-thing (literally) what has a significant impact to our life. It can improve the quality, the attractiveness of our daily life.

Thus, cars, notebooks, mobiles and garments has become sexy. It helps us feel attractive. It is sexy not only to have these things, but also show them to the public (not only neighbors and friends) and share the idea of possessing it (”it was the guy who invited me to have this”).

Web 2.0 is not only a technological trend but also developing the range of what is sexy. Nowadays, it’s much more important to be a part of the community of Facebook, LinkedIn, YouTube, StumbleUpon, Gmail etc. It’s sexy to use it. To be visible as a member, with a lot of friends.

If you would ask your friends why they’ve signed up for social networks, most of them probably did it because of being attractive among their acquaintances.

Another example is a trial to revive Netscape.com. It was tried to develop a web2.0 website under this domain. But it had to fail because Netscape is - and always will be - a symbol of web1.0, an old-fashioned and unsexy nowadays tool.

Semantic web can be sexy if it would challenge transform an obscure theory of semantics into anything what is sexy for individuals. It’s not only an aspect of usability of websites offering web3.0 (semantic) solutions. It’s a matter of being really attractive, even seducing for human beings. These services have to be not only extremely useful, they need to be attractive to share with friends.

It is a question: ‘if I showed this website to my buddy, would I be considered as a stranger or a leader of new trends, someone who’s sexy and following me would be attractive behavior?’.

Why Google won this challenge? Because using it was like showing the others: “I’m better than you, because I can find <it> much faster and easier.” Why did it Facebook? Because it was like “We’re a part of VIP community. We’re together, we can reach anybody in the world.” What about Twitter? It’s a sign of “I’m important to my friends, so I need to acknowledge them how I am”.

On the contrary, look on VOIP services. It’s growing, but it’s rather not sexy. It’s not worthy to say whole the world “I’m a miser, so I’m using VOIP”. You can tell about that a few closest friends (because you care for their finances), but that’s all.

I’m observing how are described new start-ups. Eg. Mahalo - I am not sure if it’s already sexy, but I guess that it would be, if only the idea of “human-powered search engine” would be great enough to beat Google-like SEs. Thus, it’s a case of being “human-powered” with more relevant scores, not only the fact of effectiveness (what was the main reason of being sexy by Google, when it outbeaten Altavista and others a few years ago).

I’ve recently read an interesting article about failures in start-ups and ways of managing them from a VC investor’s point of view. The most important thesis of that is quick transforming the businessplan usually is a proper move, after realizing it hasn’t been good enough. In other words: it’s better to change your way promptly as soon as you realize that your project is going to dead-end.

What’s more, this investor underline that majority of start-ups’ businessplan are flawed and need to be generally changed during implementation of the project, before getting a traction. In effect, there are much more of great successes among those companies which have changed their businessplans, than those which stuck to the plan. There is a graph being a picture of that.

In our previous projects, we’ve made a lot of mistakes which prevented us from achieving goals. We’ve assumed that the community of ZP project would have sticked to our rules, have been active thanks to only core functions and enable to use our advance system. It has failed, as well as the assumption of efficiency of a kind of innovative promotion of expert services offered on ZP. We’ve understood that it doesn’t work but quite late.

What’s more, the cost of creating our first version of the product was too high. It was so, due to the fact that we were convinced that we’ve got the traction and we need to stick to the plan and develop a great, advanced product since the beginning.

Today, I’d do it another way. Instead of spending a lot of cash for developing: first strategy, then product, I’d create a simple product based on intuition and test it.

First stage of test: talk about the idea to the friends and strangers. I know that a feedback from strangers is more reliable but it’s also much more difficult to obtain. When it comes to the friends, at least checking if they’re catching the idea and are able to show me the benefit for them from using our product, would be sufficient for a start. Then we’d try with strangers…

Second stage: show a simple demo of the product for limited number of people - only to those who would give us an overall feedback, and we’ll be able to receive it properly. It can be done as a set of individual (live-)tests aimed at performing specified tasks by alpha-users and checking if the effects are as expected. Then, if we’re sure that everything goes right, we can go to the next step. If not, it’s essential to go back and go through above points with a concept a little bit changed.

Third stage: limited/private beta version. A little bit more complexed product can be then assessed by more users, but still limited number - to make buzz, increase interest and have time to prepare for the first impression to the public.

With some projects, I suppose that the second and third stage may be tied and performed together.

In addition, I’d like to quote an author of previously mentioned article, Fred Wilson:

“My friend Dick Costolo, co-founder of FeedBurner, describes a startup as the process of going down lots of dark alleys only to find that they are dead ends. Dick describes the art of a successful deal as figuring out they are dead ends quickly and trying another and another until you find the one paved with gold.

To go back to Dick’s analogy, you can go down lots of blind alleys if the cost of doing so is low. But if you are spending a million dollars on each blind alley, you’ll be out of business in no time.

Regardless of whether you have taken venture capital or not, capital efficiency and bootstrapping are critical values. You must keep your burn rate low until you can show without a shadow of a doubt that you have a business model that works, can be operated profitably and is ready to be scaled.”

There are also some interesting Q&A and opinions among the comments:

Thomas Purves asked:

“It seems to be there is an interesting contradiction between “Most venture backed investments fail because venture capital is used to scale the business before the correct business plan is discovered. ” and the fact that 2/3rd of your successful businesses changed direction *after* being funded.

So how does this affect your investment decision. Do you go with the entrepreneur who has total confidence in a specific direction for the company, or do you do you go with the entrepreneur say with a great idea but who admits to not knowing yet the best direction (at the risk they might never find it, or just aren’t ready for VC yet).”

And the response of Fred Wilson is:

“”scale” is the key word. if you use your VC money carefully and keep your burn low until you’ve found the right model, then you can change direction after being funded and things work ok.

it’s the situations where the VC money is used to step on the gas before you really know where you are driving to that causes the problems.”

I’ve been recently investigating a buzz about TrueKnowledge and Powerset and stumbled upon an interesting history fact about some researches on pre-semantic technology developed in the 70s. As Ian Clark wrote:

“The questions selected by the person giving the demo are presumably selected such that they know the system will interpret them correctly. The real test is how well the system works with questions chosen by people not familiar with the system’s capabilities.

Back in the early 70s Terry Winograd developed a system called SHRDLU, which you could talk to in natural language and instruct it to do things in an artificial world. When used by its creator, the effect was impressive. He would type instructions in English, and they system would obey them flawlessly.

However, when users unfamiliar with the system’s capabilities tried using it, the results were invariably disappointing, because they weren’t limiting their interaction to within the system’s capabilities.”

it’s a real problem which faced Hakia. They’ve developed a kind of similar technology (maybe more like Powerset, not TrueKnowledge) and advertised themselves as a search engine which is able to answer questions like “What are the best pills for a headache?” or similiar, basic questions. They launched their service for public, despite beta-stage of their products. These are effects:

A lot of spam and useless links is a real threat for these technologies. Then, since they’ve published their demos

I’ve been reading about some intelligent search engines, developing a semantic way of organizing the web and human stores of knowledge. Recently, I’ve signed up to: twine.com, powerset.com, trueknowledge.com and checked an alpha version of freebase.com.

The three above websites are on invitation only, thus I’m waiting for it. On the contrary, Freebase.com is open to be tested by everyone, without signing up. So I’ve decided to try this idea, using FreeBase as an exemplary semantic website.

The design (I mean: usability) of this website is quite poor for me, because I needed to watch a video, showing how should I use an alpha version properly to have any accurate results. Actually, I tried to browse a directory, which turned out to be unefficient (despite it is shown in the middle of the mainpage - for what?). A video suggested me using a search engine, then I’ve done so.

I typed “polish” into a search box. I’ve been quite impressed by the results. It’s returned more than 60 different definitions of names containing a word “polish” or being highly connected semantically with these phrases. Compare with Google scores - FreeBase scores are far better.

Entering a site with a longer description of particular phrase, for example “Polish złoty” you can some see good ideas to order data from this topic, but it’s in a very early stage (but I do understand that it’s only an aplha version). However, I’d really appreciate adding some external links, which are worthy to recommend to extend this topic.

I’ve tried to do some searches of that phrase on wikipedia.org and wikiseek.com, but imho they haven’t resulted with anything better (it’s in between FreeBase and Google quality of scores).

Actually, this comparison is made with only one trial, with only one phrase - a kind of encyclopedic one, so I won’t come to a conlusion that it’s a better way to search by this website in every cases.However, in this case, I think they won.

I’m just after reading You are wasting time. Find out why about difficulties with efficient searches, not only the Internet but also other resources like corporation’s networks or even your own documents on your PC or Mac.

I insist on an opinion that there is a need to increase the power and value of whole information in the Internet, especially with making easier and more convenient access to the most proper information. To make that the information which already exist can be assessed as a HQ or a crap; can be underlined as something important and up-to-date or cross off as sth what has no use except of historical value.

I’m wondering what can be better than intelligent search engines. Some say that cross-linking can be the future. Looking into this idea, at first glance you can see that any catalogs are no longer gaining popularity - actually they’re in their agonies. However, gathering the information into one place from the whole web seems to be a good idea - think about the success Wikipedia, for example.

What about other web services, helping us to order the store of knowledge? Thinking of Digg, Del.icio.us or any search engine they’re not so relevant as they’re expected to be. There should be a good way to manage whole human knowledge in much smarter way than it used to be, or even already is - mentioning Google. In the next post, I’ll summarize what I’ve been thinking about.