The NextWeb 2008 (day one)

Although last year’s NextWeb conference had good coverage in the blogosphere, this year everything has been professionalized around the yearly event. One of these improvements is the nextweb.org, which has become a central blog site where professional bloggers keep up with the developments around new internet ventures.

You can read about all noteworthy and sometimes even anecdotal events there, so I limit myself ot some personal observations at this place.

Noteworthy was the first keynote by Adeo Ressi, “Get Funding for Your Dream“. According to him, now is the best time ever to start a new venture. But at the same time, there are many dangers luring in VC funding, which you should be aware of.

One of the most central statements: you ore strictly on your own for reviewing the contract terms when it come to closing a deal. Your legal advisor will be honest with you up to the point when you sing a contract with them, as they have just one incentive left afterwards: close the deal and get the percentage of the value you negotiated earlier. Every delay is just wate of time – so forget about honest advice on VC terms.

This reminds me of the peculiar situation we have with real estate brokers and financial advisers over here: these people all work for a percentage of the deal, so nobody is at your side when it comes to choosing the real best option, let alone a careful review of the terms.

The rest of the talk was about what to expect when going through the movements, from choosing investors, preparing your references (they will be interviewed, even the unlikely ones, and should always be unconditional positive about you) and, indeed, bad terms vs acceptable ones.

Interesting – and enlightening – it looks like we are doing pretty well regarding our own startup Twones.com.

The keynote by Leah Culver of Pownce was charming and gave most of all insight in the networking aspects of starting a online business. Her suggestion to talk more about the how and why around OAuth was not accepted by the audience. Regretful, I would have liked a quick introduction in this emerging standard as an alternative to all those proprietary solutions for all those social networks.

Nova Spivack of Twine held the keynote I was looking forward to the most. This time, surprisingly, the audience chose for an introduction into the semantic web, rather than a presentation about Twine.

And this presentation was well done. No new or surprising elements for those who follow Nova Spivack’s blog (his “CEO blog” at Radar Networks), but I am sure that many people in the audience will have “got it“. And from personal experience I know how difficult it is to explain the relevance of the highly abstract and often complex elements of the semantic web.

What I liked was the perspective in which Nova places the semweb:

Tagging approach
pro: easy to do
con: easy to do (inconsistence, no “meaning“)

Statistical approach (Google)
Pro: pure mathematical algorithms
Con: no understanding of the content

Linguistic approach
pro: true language understanding
con: computational intensive, scales badly, one domain at a time

Semantic Web approach (radar networks, dbpedia, metaweb, talis)
pro: more precise queries (metadadata)
con: lack of tools, who creates the metadata?

Artificial Intelligence approach (cycorp)
pro: this is the holy grail!
con: never finished and always outdated (the holy grail)
Now the Semantic Web approach is in the middle:
Software needs some improvement and you need metadata
But: advantages add up to a network effect; if I enhance my data, I get the benefit inr eturn that my data now can be linked automatically in all kind of related contexts, especially those I never could imagine myself.

And this is taking off at an increasing speed, see the updated graph on open, linked data on the web.

The Growing Linked Data Universe
Characteristics of the semantic web approach:

  • Make data smarter vs make software smarter
  • Metadata vs AI & linguistics
  • Open data enables network effects

Approaches:

  • Bottom up (you need to learn RDF and such) – this is not going to happen (note: basic semweb technologies exist since around 2000).
  • Top down: software builds all the RDF and OWL and stuff for you. Not surprisingly, this where Twine aims at.

Some notes on the practical side. Nova dislikes the term Semantic Web as being to vague, “Web of Data” would be more appropriate. And then, already an old theme, he adapts the popuplar but heavily overloaded term “web 2.0″ to mean “the second decade of the web” en so, web 3.0 as the third decade, roughly 2010 – 2020. So we got a timeline. And right now the early adopters are emerging, the first killer apps will be launched roughly between now and the next two years.

Finally, a critical not on business models: how do protect my business if all data has to be open and free?

The bottom line is taht every entrepreneur needs to decide for themselves, but in the long run people will move away from closed environments where they only put effort in, without being able to get the value back of their own data, let alone benefit of the network effect.

Again, this is an area were Twones will shine: our business model scale along with the network effect, the more open and the more shared each user’s data is, the more value everyone will get out of it.

Oh, and I got my private Twine invite (looks good, many thanks Nova!).

Got curious about Twones?

We will lanuch an invitation only beta at the end of the month, you can register for the beta waiting list at http://www.twones.com