Four Pillars: Of treasons, stratagems and spoils: A sideways look at search and music

The man that has no music in himself

Nor is not moved with concord of sweet sounds

Is fit for treasons, stratagems and spoils;

the motions of his spirit are as dull as night,

And affections dark as Erebus:

Let no such man be trusted.

William Shakespeare, The Merchant of Venice

It’s got to be one of my favourite quotes; it keeps coming back to me, even though I have no recollection of memorising it while at school. Powerful words. Let no such man be trusted indeed.

My epiphany about Four Pillars is very deeply rooted in listening to, seeking to understand and trying to learn about music. There’s something about music, sufficiently distinct and separate from commerce as it were, yet deeply intertwined in our daily lives. A part of me believes that there isn’t a blogger alive that does not like music; Erebus and blogging don’t go.

We can learn a lot about syndication and streaming and “publishing” from looking at how music is played or broadcast; the way we connect to music has lessons for us in search; the process we undergo in acquiring music has much to teach us about fulfilment; and much of sharing and co-creation has roots in the world of music.

It is in music that I come across the issues and problems to do with IPR and DRM; in music that I see the attempts to control access and distribution at every point in the flow, be it chipset device or connect or format or source. It is in music that I can see many dimensions: copyright of the written and symbolic forms, and of the lyrics; performance and broadcast rights; ancillary markets to do with videos and DVDs and T-shirts and coasters and ties and what-have-you; image rights and memorabilia rights; the implications of sampling and mash-ups, of creating new from many old; shattering of old distribution models and challenging of new ones; battles royale on disk formats and DRM techniques; an incredible mishmash when it comes down to the device of choice. Is that a phone a camera a PDA a music player a computer in your pocket. Or are you just pleased to see me?
It’s Yogi Berra time. When you see a fork in the road, take it.

An aside, before I get on to my theme of music and search. Have you ever wondered about how the Seven Ages of Man manifests itself in the music space, or is it just me? When I look at where I find the music I particularly like, it seems to me that there are clear changes as I grow older.

  • Age 1: Parents choose. Any colour you like as long as it’s black.
    Age 2: Radio or equivalent. Experiment and learn.
    Age 3: Latest Releases Singles. Savour your independence and spend your pocket money.
    Age 4: Mainstream Alphabetical. You’re now one of the crowd.
    Age 5: Specialist sections like Rock or Metal or Hip-Hop or Classical or Reggae or Folk. Now you need to work at looking cool.
    Age 6: Easy Listening. Count your teeth and check your hair loss.
    Age 7: Between the candy and the chewing gum at gas stations and convenience stores. Book that funeral parlour.

On to the meat. I was reading last week’s New Scientist (I was travelling when it came out), and found this article where Kurt Kleiner argues that it’s time for a “whole new kind of search engine“. Unfortunately the link only gets you to a stub, the full article has been DRMed out of the brownies. Sad.

His basic thesis is this: There’s an incredible amount of music out there in the digital universe, over 25 million tracks. [Just think: If you laid them all end to end you may actually surpass the queue consisting of unemployed (?) Sarbanes-Oxley consultants. Unless they return to full employment as Digital Rights consultants….] Traditional indexing methods aren’t good enough. Even tagging is not good enough. So it’s time to create a more elaborate way of describing a tune, taking into account “not just key or tempo, but dozens of different characteristics, including the timbre of the sounds, chord progressions, the individual instruments and even details about each singer’s voice”.

He mentions what MusicIP seeks to do in this space, including the MusicIP Mixer as a product, and the company’s plans to develop “software that uses information on your lifestyle, diet and tastes to come up with meal suggestions, recipes and even a shopping list of ingredients”. La la la I’m not listening.

Mr Kleiner also covers what Pandora does, reporting on what Tim Westergren, the founder, has to say: “He calls the system the Music Genome Project: Just as individuals can be identified by their different combinations of genes, so Pandora aims to distinguish any piece of music according to how it scores on this set of musical “genes”. ”

Try and read the article if you can, it’s worth it. Just to put some new stuff in your head, or to shake some old stuff up. But in the meantime….

It made me think. There’s a lot of activity in the music space already…..

First, old-style deterministic search. Exact matches only and all that jazz.

In this space, Shazam already works wonders for me, I find it liberating to point a phone at the source of music while in a car, and to have the details of the song that’s playing texted to me. Shazam’s been around in the UK for a number of years now, and it’s pretty good. [Killed off pub music quizzes though…] So someone somewhere has done some work on the technique of converting music into unique digital patterns, and then yanking exact matches out of a large database. And yes, the Pandoras of this world can improve on the model.

Next, more modern probabilistic search. No longer exact matches but analogies and parallels, relevance and ranking.
I remember hearing about MongoMusic some six or seven years ago, they were the first firm I knew of that had an answer to the “Sounds Like” problem. But before I could get to play with them, they were taken over by Microsoft and disappeared off the face of the earth. I have no idea whether they will resurface in Son-Of-Windows-Media-Centre-Meets-Vista. But they could.

Then, collaborative filtering approaches to music.

While Pandora does its bit (and I am sure there are many others) last.fm seems to have this market well in sight. People who liked this also liked. Share playlists with people like you. And all that jazz. Or classical, if you prefer. Which reminds me, Firefly were the first guys I saw with a real understanding of collaborative filtering. And they too got Redmonded. And disappeared. Another to resurface post Vista? They could. Firefly founders seem to have re-established themselves in Skype and in Lovefilm, amongst others, so they’re obviously still in the same space.

We can then move to better visualisation techniques involving music.

The best I’ve seen so far is what was called MusicPlasma, now LivePlasma. Again, there are many others, all I am doing is declaring the one I am most familiar with. Here we have some sort of almost-fractal images to depict artists and groups and genres, a blueprint for finding “neighbours” of musicians you like.

And just in case you need a WayBack Machine for the rich journalism that underpins all this, you even have labours of love like RocksBackPages; again, I’m sure there are others.

The examples I’ve chosen are illustrative and no more than that. The list is neither comprehensive nor necessarily accurate. And I’ve avoided the simple established ways of doing things like iTunes. Which is great. For now.
What I was seeking to do is to prove a point.

Modern search will consist of a number of things: exact matches when called for; Sounds-Like when appropriate; collaboratively-filtered when appropriate; playlist-traded when possible, on community or friend recommendation; working off graphical visualisations when appropriate, a GoogleEarth-meets-LivePlasma approach; driven by rich history in text where relevant, a Rolling Stone meets RocksBackPages approach. All surrounded by these places we call the internet, with a host of little markets like the ones described above, and a host of little communities ranging from eBay thru Amazon thru technorati to YouTube and Bebo.

That’s what all search will become, when Generation M rule. And our role, from an enterprise-meets-technology perspective, is to pave the way for them. By doing the right thing with operating systems and platforms and infrastructure (keeping them vendor-agnostic and affordably priced), with devices and connect mechanisms (keeping them diverse and versatile and completely at consumer choice level), with digital information (avoiding all the dinosaur pitfalls in the IPR-meets-DRM space), with personal and collective tagging.

That’s why I try and learn from music when I want to build out the Four Pillars model. Because people are working on it now. Experimenting now. And it’s exciting.

BTW something else occurs to me. I’m glad someone didn’t patent each and every note in music, because I’m sure it’s not for want of trying. People will try and patent the strangest things. But if we had some sort of genome sequence for each and every song made, wouldn’t it be fun to watch someone trying to stop sampling and mashups? Sorry, it’s now a new music DNA segment, it’s not a copy. You live by the IPR-DRM sword, you could die by it. And the sword’s pretty blunt. Which is why we need a whole new way of compensating artists, rather than paving the cowpaths of the old regime, as Michael Hammer would say.

Four Pillars: Of markets and conversations

It’s been an interesting day.

Marshall Kirkpatrick’s blog gave me some views on what Google and Feedburner are up to. Google Video for Mac. The deal between Feedburner and Geffen.

I got to BL Ochman’s blog via David Terrar to find this story about eBay. One of the world’s biggest marketplaces taking blogs and wikis seriously. Maybe markets will become conversations in the digital world as well.

Sean, time to dust off that AmazonBay video.

Four Pillars: On trust and sharing

Coding The Markets, via a comment on one of my recent posts, pointed me at Larry Harris’s book and associated website called Trading And Exchanges. As expected I’ve ordered the book and look forward to reading it.

Harris makes a nice distinction between trustworthiness and creditworthiness.

Something still niggles me though, above and beyond what I stated in my reply to the comment.

It’s about trust and sharing. P2P models of creation and co-creation work on P2P models of trust, and sharing is simpler as a result. When trust becomes something administered by a central body, most of the time it morphs into an unworkable administrative nightmare of hierarchical and structured and graduated permissions. A nightmare that becomes closer to Treacherous rather than Trusted, to borrow a phrase from the infosec world.

I just can’t see a P2P world buying into that kind of structure. Trust has to be something that is part of a relationship and simple and yes-no. I hope we never need to implement hierarchical need-to-know and eyes-only and clearance-level based approaches, because they have unintended consequences. Or at least unwished-for consequences, speaking personally.

They limit collaboration.

We have to find other ways of building trust models. And the permissions we give implicitly and explicitly as a result. I must look into powers of attorney. It is possible that they had well-bounded scopes, but I cannot believe that lawyers would have come up with a graduated model for the instrument; not when the instrument was an integral part of an attorney-client relationship.

I must look closer at Zopa and at GrameenBank as well, to try and understand what happens in the P2P world from a pure trust perspective. While I know them and understand their basic premises and models, I have not looked too deeply at their gubbins.

Four Pillars: More on trust and confidence

First of all, thanks for the comments and the e-mails after my last two posts. It’s good to see the snowballs rolling, gathering momentum.

Why am I doing this? Not because I’ve suddenly become some male-menopausal closet philosopher. But because trust and relationships and identity are integral components of any market, particularly financial markets. Software has caught up, it is now possible to embody some of these principles in software. It is also possible to get them hopelessly wrong, and to throw away the advantages and opportunities that technology, more particularly information technology, provides us.

Aside 1: One of Doc Searls‘ themes at reboot was that blogging is provisional. This is important. These are conversations and discussions, not dogma. Mind Wide Open, as Steven Johnson would say, but not in this context specifically. [And Happy Birthday one day late, Steven, if you ever get around to reading stuff like this]
I’ve been thinking more about why I feel it’s so important to distinguish between trust and confidence, why I am deeply uncomfortable with the definitions getting intertwined. Language is something that’s living, and meanings and nuances change over time. Years ago, I was fascinated to learn that “nice” used to mean “complicated”, as in “That’s a very nice two-move chess puzzle”; that “fond” meant “foolish”, and conjured up images of fresh-faced Wodehousian wonders mooning cheerily over absent first loves. More recently, we’ve had the Governors of the BBC pass judgment on the use and/or abuse of the word “gay”, to Chris Moyles’ consternation and relief.

I remember the first cricket match I ever went to. And what a match; details to be found here. Gary Sobers, Conrad Hunte, Rohan Kanhai, Lance Gibbs, Clive Lloyd, Seymour Nurse, Wes Hall and Charlie Griffiths, facing off against the one-eyed Nawab of Pataudi, Rusi Surti, Chandu Borde, ML Jaisimha, Abbas Ali Baig. Bedi playing his first Test. Chandrasekhar and Venkataraghavan playing as well. Each captain bestriding the field like a Colossus. [A confession. I first tried Colossi, didn’t like it, it didn’t work. Changed the structure of the sentence. Reminds me of the Alipore Zoo curator rumoured to have written to his counterpart at the Sydney Zoo, asking for the despatch of one mongoose. And then asking if he could send a second one at the same time :-) )

India lost. But no matter, it was a great spectacle. With a terrible start. A dispute about an umpiring call led to fans in one section of the ground pulling up the seats and setting them on fire. On New Year’s Day. Setting the backdrop for the newspaper headline “Hell at Eden” the next day.

Which brings me to the point of the story. I was nine. We had to evacuate the stadium in a hurry. There were flames surrounding us. And the only way to get out was to go up, and then jump down. My father shepherded me to the jump-off point, told me that he would jump first, then catch me when I jumped.

I had no confidence that my father could jump down 20 feet without doing himself a major injury; it would have been the most energetic thing I’d ever seen him do, other than hit a golf ball. But somehow he jumped, and landed with limbs intact, aided and abetted by the throng below. [I’m not sure he could have jumped anywhere in Calcutta without landing in a throng, but that’s a different matter.]

I had no confidence that my father would catch me. But I trusted him. And catch me he did.

Trust is binary. Yes or no. Trust is sacrificial, you have to make yourself vulnerable in order to trust. Trust is two-way; A cannot trust B unless it is also true that B trusts A. Trust is a relationship, a bond. It has something to do with emotional intelligence and faith and Pay It Forward. It is based on something holistic to do with the person and the relationship.
Confidence is graduated, it is part of a continuum of values. You associate confidence with words like ratings and levels and “degrees of”. It is one-way, if A has confidence in B then it does not follow that B has confidence in A. It is an attribute about a specific skill or ability.

Trust is Because Of. Confidence is With.

Aside 2: My grandfather was a Founder Member of the National Cricket Club at Eden Gardens, founded before my father was born. Founder Members were entitled to two pavilion tickets, and you didn’t even have to wear breakfast-coloured ties to sit there. Founder Members were entitled to these two pavilion tickets each for life. Between 1967 and 1980 I watched pretty much all five days of every match at Eden Gardens. Using Founder Member tickets. The youngest a Founder Member could have been, in 1969, was 63. I was 11 then, and not the youngest queueing up either. And nobody said a word.

This is an example of what’s happening today. One of the things the Web is challenging is the market lock-ins that prevent natural secondary trading from taking place. More and more, we will see “not-transferable” items becoming more liquid, more tradable. Entertainment and sports event tickets are the likeliest candidates to start with, but even airline tickets will go that way.

To make this happen, we will need better understanding of how identity is altered on such instruments. The answer is not to prevent the transfer, but to solve the identity transfer issue. There will also be cases where the “instrument” is transferable only after a “first refusal” hoop-to-be-climbed; this is reasonable as well. We have to solve the process by which we can portray identity accurately and allow for first-refusal clauses, rather than prevent something as natural as secondary-market trading.

Four Pillars: Confidence and trust

Martin Geddes, while commenting on my previous post, threw a snowball at me. He recommended that anyone interested in understanding the difference between confidence and trust should read Adam Seligman’s The Problem of Trust.

I trust Martin.

So I ordered the book straightaway, and Amazon have guaranteed that they will deliver the book to me by 1pm Thursday.

I have confidence that Amazon will deliver.

I tried to figure out what I really mean by these terms, in preparation for reading the book. And what I came to was this:

I believe trust is about what a person is, about people, about relationships. It is integral and whole within itself, and is based on values and ethics. About covenant.

Trust is vulnerable. It is like faith.  You believe someone will not do you harm.

I believe confidence is about what a person can do, about his or her abilities and skills. It can be segmented into degrees and levels, and is based on transactions. About contract.

Confidence is set against past achievements, and starts at zero and builds from there. Trust is set against beliefs and values, starts at 100 and builds from there.

I have no confidence in Martin’s ability to predict the finalists in this year’s World Cup. I have no confidence in Amazon’s ability to predict the finalists in this year’s World Cup.

But I trust Martin’s recommendations related to books on trust, and I have confidence in Amazon’s ability to deliver said book when they said they would.

Comments anyone? I will try and expand on this after reading the book. With issues like this, I find I learn faster if I articulate what I think I believe in the first instance. Even if all that is achieved is that I know a little bit more about what I think.

What I have written so far on identity and trust and relationships and authentication and permissioning is based on what I have stated above. As I learn more about these things, I’m sure I will change what I have written.

 

Â