Montreal Tech Watch

There were a half-dozen keynote speeches yesterday, sandwiched between the day’s seminars, exhibits, and the sideshows of the unconference, and the evening’s libations and mixers around the offices of San Francisco notables.

Tim O’Reilly went on stage to repeat what Bob Metcalfe and others were saying over 12 years ago, that the network is really the computer. Tacked on were two side concepts. The first idea is an invitation to tackle large common-good projects so that even failing is contributive. The second is an interesting take on the market’s valuation of centralization (Facebook, Google, etc.) even as Web 2.0 is pulling the web towards decentralization (Open APIs, shared contexts, etc.). The end-result is that market-valued centralization will happen through interoperability. The unstated conclusions are interesting though; we can’t value or buy a share in inter-operating companies, unless through a mutual fund (assuming the companies are public) or a Yahoo-style consolidation (if not). Is centralization dressed in new clothes still the same old successful maid of yore ?

The most expected talk of the day was the announcement of Microsoft Live Mesh, a long-haul project built and hyped under the supervision of Chief Software Architect Ray Ozzie. Once you take all of the buzzwords about collaboration and data synchronization out, you essentially get, as far as I can tell, a RSS-enabled shared folder with a public changelog and a programmable API. The first application of Live Mesh is one in which multiple devices can share preference and settings files (bookmarks, contacts, personal Windows settings, the kind of thing you get for a decreasing premium on certain USB keys) so that they are all using the same basic data (as long as they run Windows, although Microsoft promised wider support to a snickering audience). They claim over a hundred developers were assigned to work on this for two years. Taking into account the complexity of building shared-storage systems (instead of, say, collaborating with Amazon or acquiring the likes of Nirvanix), I wonder what the other 90 were doing.

By far the most interesting talk of the evening was a live stage interview with Max Levchin (PayPal, Slide). If you are running low on smart, well-articulated, incisive content, you can always count on Levchin to deliver. Max covered his early attempts at start-ups (4 of them until he found success with PayPal), but focused especially on the social entertainment software that is the core of Slide’s applet business. He went at length exploring the relationship between social actions and advertisers as a non-abrasive promotional vehicle; witness, for example, the addition of a wildly popular pregnancy test to be thrown at others in SuperPoke to coincide with the release of the movie Juno. He covered new ways to segment the market based on behavioral commonalities rather than demographics, an idea that the market analysts at an earlier Consumer 2.0 panel hinted at. Levchin then offered an interesting distinction between applets and traditional software: that applets draw on users’ wish to participate through one destination, made valuable through its character and popularity, unlike traditional applications which are meant to be chosen not for their intrinsic identity but rather for the predominance of certain features and qualities differentiating them from the feature lists of others; that this is what makes widget companies so valuable. Interspersed in the talk was a fourth idea on the lifecycle-prolonging value of widgets as the novelty of social networks erode. Good stuff.