Making stuff as a founder of Avocado. Former music-maker. Tuna melt advocate. Started Google Reader. (But smarter people made it great.)

A Rambling Speech to the Fictitious "Web Art Society"

First Draft, Speech

Hi, thanks for asking me to speak today. I'd like to take this opportunity to acknowledge the wonderful work accomplished by many of the people we see in attendance here today and I feel honored to be among the assemblage. It's gratifying and humbling to witness all of this combined brilliance and entrepreneurship struggle to park their SUVs into scooter slots.

[Wait for laughs. They might last a while.]

My thesis today is about change. The change we encounter in a global sense. And I'd like to be critical about it, if I may.

[Clear throat. Begin slowly.]

Evolution, like Internet Explorer, is broken.

[Pause for shifting, scorning sounds. Foo? Snarf? Shizvf?]

Lemme explain. One problem I see with Plain-Old-Biological-Evolution (and the currently oft-maligned IE) is that both systems have a rate of adaptation that is well, evolutionary. I mean, they update slowly, people. Evolution, in particular, doesn't like to keep a snappy pace; it lacks a serious sense of urgency, and we'd be reasonable to conclude -well- that nature doesn't really have a good work ethic. It's not so much "in the zone." Ol' Evo just doesn't want it as bad. In fact, it's the laziest set of processes I can think of other than long-term atomic decay or universal expansion (current day, that is).

Some would counter, reasonably, that nothing good changes quickly. Yeah. Ok. But, seriously, no. The relatively rapid pace of technological advancements and the value they bring are handing tired ol' Evolution its ass.

[Gesticulate and punctuate a fervid "In!" "Your!" "Face!"]

What's wrong here? To put it simply:

Evolution doesn't learn.

There are processes out there right now whose approach might help inform improvements to Evolution. But Evolution is static and closed-source, an altogether unforgiving combination - as you former Microsoft competitors may well know. [Wait for over-smug laughs and snickers.] But, hellooo, Evolution teaches us the value of adaptation and it follows that re-evaluation of one's scientific approach is self-evidently efficient - so metaphorically speaking, why can't Evolution drink from the same mug it put in the cupboard?

Evolution: An Unimaginative Parent

Let's put a fine point on it. There are concepts basic to millions of parents that are completely foreign to Mother and Father Nature. Evolution's strictly not interested in growing and nurturing the past. What's its apparent mission statement? Trash everything and make new stuff constantly. And it does - by consuming bodies by the googol and passing genes like a runner's torch. The emphasis is all on up-front cost and much less on planning and saving.

Evolution plays the game like a 1999 VP of Business Dev.

Mortality is Evolution's big shortcut. Burn resources fast, maximize value, and make disposable skins. Take the hit on replication but improve adaptation through massive (and distributed!) trial-and-error. And I can appreciate the choice being made, Evolution's reliance on Mortality is just making the best out of bad situation. If you're handed a menu where every dish includes Entropy as an ingredient, well, you just have to eat it.

[Watch carefully for delayed laughs due to slow realizations of juxtaposition of idiom and metaphor.]

But Evolution could substantitally benefit from reading a sysadmin cert's braindump. For those who aren't familiar with the term braindump, it's geek jargon for "cheat sheet." And renaming this term is a geek concept called "Security Through Obscurity". [Laugh first so that others may follow.]

Now I mention sysadmins for a specific reason. It seems clear that Evolution just doesn't know what to do with Humanity and that a junior level sysadmin, well, would. Evolution seems unaware of how to leverage Humanity's best two assets:
  1. Consciousness
  2. Human-Level Invention
But, ooh, ooh! [Raise hand like excited student.] We know how to deal with them. At least we think we do. And given full exposure to the underlying source code of life we should feel confident that humanity could improve the accuracy rate of adaptation overnight! Or at least before Longhorn ships.

[Wait for laughs, not too long though, you've hammered this MS thing 3 times.]

In fact, I feel that a sysadmin's daily grind solves the kinds of problems that evolution-via-massive-doses-of-mortality currently ignores. In at least three areas...

If a segment fails (and I mean really, if it "dies") then an entire lifetime of accumulated knowledge which otherwise could be fed back into a useful QA loop is immediately lost. And if the segment-in-question hasn't procreated, successful adaptations that were unrelated to procreation are lost.

This is nasty. I mean, just try and transfer data from one human to another. You know what it is? It's sneakernet! You have to form words and deliver them across the ether. And complicated recombination and interpretation algorithms are required in any humans engaged in transfer where, often, the algorithms are independently maintained and differ in their interpretations. To further complicate things the transfer medium is, get this, the thin and easily disturbed wavespaces of light and sound. Completely nuts.

The body's genetic information is stored at a very low level in a layer for which there exists no native monitoring tools to help collect and measure important adaptations as they happen. Adaptations have to be passed on to survive and not all important adaptations are related to mere individual survival. Also, all consciousness is stored in memory which ... could ... be fine but data loss is tremendous since (this is delicious) the buffer is shallow. There's only so much memory a conscious being can accumulate. Certain memories are discarded or buried by some arcane algorithm which attempts to jettison, distort, or bury memories it considers unimportant. Apparently, its OK to occasionally forget whether or not you're drinking from a pond where weeks ago you saw a dead animal submerge and decompose. "Oh man," the deer said, "that was this pond?"

Culture: A Workaround to Some Evolutionary Bugs

But any well-framed criticisms we have of Evolution are merely silence to the Universe's deaf ear. There are no appeals, no amendments. Evolution is oafish - "I'm not listening to you, I'm busy!" So, like any group of effective sysadmins, Humanity has workarounds to the obtuseness of this particular, selfish co-worker.

We have Culture, for one. Culture is a data set and a technique that uses some pretty nifty things (little things like Language, Print, Images, Speech, and Music) to help store accumulated knowledge and pass it along to new Humans. This helps slightly with the problems I've labeled earlier as Failover and Storage, though the sheer amount of information available for storage and the problem of our buggy light-and-sound-based Throughput makes importing to any store an awkward and constrained task - a problem that has, in evolutionary terms, only been recently exposed. Information is often Lost in Translation, indeed. For example, does a video recording of a human eating a grapefruit also store their inner thoughts or their unexpressed emotions or the metabolic changes? No, I say. It does not. Not on VHS, at least. Though I presume TiVo's working on it.

[No one will laugh at any of this, move quickly.]

So Culture, too, has a few bugs. Which is important to note since bug reporting doesn't impress Evolution. Remember? Deaf. Snobby little shit won't help us with this problem now. Maybe in a few million years we'll have an external appendage which stores and shares memory or maybe we'll have perfect recall via improved synapses. But so many millions of terabytes will have been generated, logged, and destroyed by then that the solution will be successful only on a pyrrhic level. 'Cause also, hey: The Sun. Gone. Sucks to be us - Evolution works slow, is buggy, and as far as anyone can tell, doesn't know Astronomy.

So we meticulously document some thoughts, some knowledge, some understanding. But our solutions for getting this information ready for analysis and dissemination to new Humans are somewhat crude. We often transfer the information to a physical medium via a mechanical interface and then ... well ...

[Clap hands and shrug.]

That's it. That's the process. Analysis occurs differently depending on the storage medium and transfer method back to the ultimate arbiters of the data, which is just us Humans. Evolution can't analyze the fruits of Human discovery; Evolution is still deaf, still mute. So we really are the captains of this fateful ship. And we steer it using Culture - which is not a perfect solution, but it's a passable workaround.

But if we could improve this hack, this workaround-called-Culture even a little bit, the value-add might be global. And that would be something.

So, let's try. In fact, here's a new problem to chew on. The (again, relatively new) amount of information available for transfer makes data analysis another difficult and immense task that is proving to be more difficult every day. Sure, I know a few people trying to solve this problem. But you know, I suspect all of us here tonight are trying to solve it, though perhaps not all of us consciously. We talk with family, friends, and loved ones, watch TV, listen to music, and generally sift all of this stuff as best we can. We chunk our thoughts into data sets with varying degrees of structure, even moral order, and decide what information to disseminate.

And then we forget. No matter what age, we forget. Randomly. We might forget the feeling we had when we first fell in love. The name of our fourth pet. How we got that scar. And more abstractions. Our first disappointment. Our reaction to mortality. An unexpected hope. A private, personal revelation. A publicly useful discovery. A method for avoiding or entering into violent conflict. A last kiss. A solution to Fermat's theorem. And, more personally important, where in the world I left those damn keys!

We forget many things. And we lose huge slabs of data upon which might be etched the exact adaptation we need to survive. Or the adaptation required to improve our shared lot, including perhaps the power to be better agents of adaptation than our genes. (Take that, Evolution, you knock-kneed, four-eyed wuss!)

So, I for one, have arrived at what might be some useful initial conclusions:
1. Culture is an important amendment to Evolution whose creation and analysis requires constant improvement.

2. Cultural analysis is much less useful if we aren't searching all of the data available.

3. It is impossible to anticipate where useful leaps in adaptation via knowledge will arrive.

4. No experiential data is unimportant. Some data is relatively unimportant - but we can't determine that without analysis.
Ok, sure, analysis. Easy to identify as a goal, hard to achieve. So, let's look at it. An important step to generating analysis is filtering data usefully. Do we have any useful filters for global data currently available? Are they powerful? Are they accessible? Are they capable themselves of adaptation? Do they have angel funding and, if not, can I get their cell number from you?

[Wait for laughs and/or groans.]

We do have some filters. And we have this idea of distributed, accessible storage that's formed as the Internet. So it's a matter of tweaking it to become a better storage medium and a better analytic tool. And it's my presumption (and hope) that this will be increasingly useful to cultural transfer.

But to be useful it requires data. Mountains of it because, as I stated before, I believe that no experiential data is unimportant. Even mis-analyzed data and outright deceptions are important, so long as corrected, verified, and distributed data exists in even greater numbers. Wrong data is important because we always need to know where we falter and fail and compromise. (Not surprisingly, I believe most spam is not important to this evaluation or has a relative importance of near zero.) So this data needs to get into storage. To the Internet, in this case.

So how do we get data there? Right now, we tend to copy it digitally from offline sources around the world.

You know, just to spitball here, I think we could improve the process by having data entered directly into storage. Directly to the net. And further streamline certain data imports through automation. I'm definitely late to class on this school of thought, obviously. For example, my cell phone copies the number of voicemails I have to a database somewhere that's accessible online. The amount of devices that perform this automation increases every day in opposition to the reasonable consternation of some privacy authorities and critics.

But a good deal of knowledge is trapped in our funny-looking heads. Lots of cool stuff. Important personal reasoning and experiences, for example. And if we wanted to increase the scope and depth of our search and analysis, people would have to add stuff. And to encourage offloading stuff and make that transfer of Stuff-From-my-Head-to-the-Net convenient for humans on those occasions where manual entry is needed we'd have to make the data import process easy, obvious, powerful, and accessible to everyone.

Yep. As the song went, it turns out that We really Are The World. [Sing Bob Dylan's verse here, unless throat is dry.] And until Evolution gets its act together, we are the universe's best source of data about human knowledge and experience. We are the stuff that could be home to that next important adaptation. And that adaptation might be in you now or it might be out there already. And it might need exposure or patronage or be voted on or flagged as important or be pushed to the front of the line in order to become evident or fully-realized. And as far as the Internet is currently constructed and maintained, every vote counts. Just ask search engines! All of us should approach this concept seriously, for whatever we link helps to maintain the state and accuracy of powerful filters. For everyone. Every link is counted. The algorithms of search engines contain the seeds for an unusual, global, and democratic process.

[Cue national anthem. Of Canada.]

But the price of democracy, as they say, is stiff beans. It requires vigilance. It requires participation. And participation online (for those who can afford it and barring the practices of certain less-free sovereignties) is optional. You don't ever have to vote, or offer an opinion or a thought. You can just read, watch or listen to things and you won't have to generate stuff for the online world, which is a reasonable expectation - but then, in my opinion, if you don't offer some of your brain, you don't get to complain about the usefulness of the filter, either. And many will want to complain, if they're able. Especially if they can't find stuff that relates deeply to their experience, their life, or their reality. The deep search will always be harder and potentially more valuable. The shallow search is easy, for example, I can find out what new movie John Woo is directing in less than .06 seconds. It's a new He-Man movie. But if no one, especially Mr. Woo, posts their opinion, it will be harder to find out why. (In this particular example, I'm worried that we may never get a satisfactory answer.)

[Do not cry. Do not cry. Do not cry.]

Or to put it another way. Don't only ask what the Internet can do for you. Most of you are in a position to know exactly what it can do. Why not instead ask yourselves what you can do for the Internet?

[Contain giggle to self for obvious dramatic miss.]

Right, yes, that question seems ridiculous in the face of global discord and personal challenges. But you're here today. You're part of the solution. Otherwise you wouldn't be here. I believe you can add your voice to the discussion. I know I should try. I kinda believe you should as well.

And that's all I have to say. I'd like to again thank all of you for the invitation and for the largely polite way you held yourselves while I droned on. I'd also like it if I could say that improvement of the efficiency of Evolution was the big, secret plan behind everything we've been attempting to create. But, hey, it's not, really. Because, as far as our access logs detail - well -

It seems it's actually your plan. It's her and it's his plan. It's, quite evidently and seemingly unstoppably, our plan. Because every day without end (and by the millions) we add to our shiny new super-mind as if our lives depended on it.
posted at October 18, 2004, 11:05 AM


  • At 6:40 AM, chuck said…

    Your fourth pet was Good Fortune, a small, black rabbit with a white stripe that lived in the garage on Briarcliff until one day when she left us, after Eedle Ortle Eedle, Tiddly, and Snoopy and though your data banks were advanced compared to other carbon units, I'm sure much has been trashed though possibly retrievable.
    Evolution and God were sitting in front of their newly purchased hardware with the latest high speed wireless connection to the cosmos. As a flash (otherwise known as blip) came and disappeared across the screen at a speed greater than light. God asked "What was that?" "Oh, nothing." responded Evolution "trashible data from a wrongheaded experiment regarding carbon units ... just a blip on our screen."
    Never expect others to be as excited or as involved with your new toys as you are (ideas are your toys of choice now Son), especially evolution. Also never pause for clapping or laughs, though deserved. It's not cool anymore to be perceived as needy. Also screenplay speeches should place behaviors other than speaking as a sidebar. "Days are long... life is short" Stephen Usdin, New York lawyer. I would add " unless of course you are God or Evolution." Evolution has the patience of a woman scorned anticipating revenge, she will outlast you and never regret the wasted energy.
    The first draft is a well crafted work of art and most likely shouldn't be tampered. I most certainly don't get it all, but who does? If you find her, marry her. I found her and wasn't ready. I'll write a short epistle soon, ex-post-facto. I will title it 'Portland Girl'. No, not that Portland.
    I'm not proud of you Son, I'm proud for you.

  • At 3:17 AM, Chris Wetherell said…



Post a Comment