[FoRK] A Theory of Products: Magic, Alchemy, Science... and Beyond?
pj at place.org
Mon Feb 8 19:31:59 PST 2010
An interesting perspective, as always. I agree, though I think it means
that no product can ever be finished because it will need to adjust to
the changes in a user's tastes over time, and it's probably so optimized
to only be 'in style' for a short time. A kind of usability overfit,
to make an analogy to machine learning. Marketers may think that's a
feature. I think it's a sustainability bug, personally, but YMMV.
On 01/29/2010 08:48 PM, Jeff Bone wrote:
> The use of the term "magical" in Apple's description of the iPad has
> been nagging at me... after quite a bit of consideration, I've come
> to the conclusion that their use of this term is "meta-anti-ironic."
> By this, I mean the following: they are attempting to use the term
> figuratively to describe one thing --- the product, which that term
> does not obviously literally describe --- but in doing so have
> unintentionally used the term to literally describe something else
> *about* the thing they intended to describe --- the mindset in which
> and process by which the product was conceived, defined, and created.
> In doing so they've revealed what I believe to be an atypical (but
> probably occasionally unavoidable; magic is, as we all know,
> unreliable) result of the way they approach product creation: confusion.
> This gives me an opportunity to dwell (at length ;-) on a topic I've
> long been fascinated (obsessed?) with: the intertwined process(es) of
> product conception, definition, creation, adoption, and refinement /
> evolution. I.e., a theory of products and a sort of mini-taxonomy of
> the evolution of how we approach these processes in making high
> technology. Unfortunately this little write-up doesn't really provide
> any such theory, despite the title. But consider it a random walk
> through some of the prerequisites for understanding any such thing ---
> which, I ultimately conclude, may not in any case be possible,
> necessary, or desirable. Maybe. ;-)
> First, definition of a couple of terms and some scoping. By "product"
> I generally mean any technological artifact; though most of the cases
> and examples considered and given are software and / or computing or
> consumer electronic hardware (fuzzy lines, there) the general ideas
> should apply to any sort of intentional, purposeful artifact --- i.e.,
> all "products" are technological. By "product management" I mean any
> process of defining products as a part of the process of creating
> them. By "product marketing" I generally mean the process of somehow
> gathering the information necessary for the definition of products.
> It should be interpreted as an *inbound*, intelligence-gathering
> process --- and not confused with "outbound" marketing such as
> marcomm, pre- and post-sales support, channel / pipeline creation and
> support, advertising, etc. Specifically this "product marketing" can
> entail any number of distinct activities, some of which may be
> described below, but which are all intended to gather intelligence to
> inform product definition decisions, requirements creation (or
> exclusion), and so on.
> In The Beginning...
> -- The Magicians --
> In The Beginning we had product magicians. Their product "magic" was
> / is a process primarily of introspection; they reflect upon their
> own attitudes about, uses of, and desires for certain technologies
> and, from this reflection, can (or attempt to) extrapolate the needs
> or desires of others and synthesize product definitions accordingly.
> The result is something equally magical: a "vision." It is gestalt,
> genius, artistry. It is aesthetic, soft, egoistic, passionate. The
> process is highly subjective, intuition driven, and qualitative ---
> and success or failure relies entirely upon the individual magician's
> ability to execute this mysterious, internal, creative / synthetic /
> syncretic process that they themselves probably cannot articulate,
> much less teach others. Success or failure also requires the ability
> of the magician to manage the efforts --- often dictatorially --- of
> others in actually bringing the product to fruition while maintaining
> the integrity of the original "vision."
> Most early successful higher technologies relied upon creators that
> were magicians and processes that were magical. Some few were really,
> really good at it, and still manage to practice it effectively today.
> Examples: two should suffice, though there are many --- the early
> computer entrepreneurial landscape was littered with them, both
> successful and not so. The master magician of them all, clearly:
> Jobs. His apprentice Jonathan Ive, is no less the true believer and
> practitioner, with his wide-eyes (from staring into the flames of
> Platonic truth) and his high temple of creative conjuration, his lab.
> (Be very quiet when you enter, lest you disturb some intricate
> spell-in-progress.) The previous heir-apparent, Tony Fadell, is
> probably an example practitioner of the next step in the evolution of
> product marketing and management towards science: an alchemist.
> (Speculation: the fundamental conflict between attitudes toward magic
> vs. alchemy is the reason that Ive is the golden boy while Fadell got
> Digression: today's "web designers" are most --- perhaps all ---
> latter-day, would-be magicians, though mostly of the low-end hedge
> wizard variety, masters of minutia, scam artists mostly ---
> prestidigitators. At their best, they may be envious alchemical
> aspirants, Their major achievements tend to be the conjurations of
> illusory mountains from trivial molehills. Oh, how they peddle their
> mysterious arts! How they dazzle and confound with their bizarre
> utterances, their mercurial outbursts, their mystical convictions and
> religious, inviolable heuristics about font, color, contrast, and oh
> yes, whitespace; layout, and (of course) the proper size,
> orientation, and opacity of drop shadows --- and please, let us make
> sure that the radius of curvature of those button corners preserves
> the appropriate ratio of earth, wind, fire and water. How they seek
> and seek that deific ideal of "user experience." They tend to
> practice a folkloric art, with some apocryphal writings (mostly grad
> student papers from the mid-80s MIT Media Lab) and oh yes, that Bible
> of theirs (Apple has a role here) --- the Apple Human Interface
> Guidelines, or perhaps their Satanic Bible by Raskin, the LaVey of
> their dark-ish arts, or (for the more sophisticated and sophistical
> among them) their John Dee: Tufte. More misdirection than magic,
> even less science than alchemy, with few pretenses towards it...
> -- The Alchemists --
> The product alchemists are product magicians that would be product
> scientists, if merely they understood how. They attempt to apply
> various external, reproducible, objective laws, observations,
> measurements, methods, and so forth to the product creation process in
> order to achieve the optimal result. However, there's still a kind of
> animistic, ad hoc, magical quality to the effort. They tend to
> reflect on the objects / artifacts themselves, their abstract purpose
> and uses (use cases, etc.) and subjective musings about how
> individuals might use any given artifact with any given configuration
> of properties in some context to some end. The methods used tend to
> be a mix of qualitative and quantitative. The alchemists have the
> right goal in mind, but generally not the right tools (i.e.,
> developed-enough models, relying instead on ratios of bilious and
> phlegmatic humors) or methods (obsessive focus on quantitative means
> of taking actual input data and turning it into objective meaning.)
> The input is too selective, the data sets too small, the processes too
> ad hoc, the experiments too uncontrolled, and the objects of
> consideration too abstract and animistic to really call what they do
> science. To a large extent the success or failure of the effort still
> devolves to the quality of the intuition of the individual alchemists
> General Magic (ironically...?) was an alchemy shop. For all their
> usability studies, rapid prototyping, quantification, feedback
> processes and loops with partners, focus groups, their UI committee,
> etc... they still managed to produce a beautiful egg-laying milk pig
> (i.e., attempted to be all things to all people, ultimately satisfying
> none.) By contrast another alchemist --- Jeff Hawkins over at Palm at
> the same time --- managed to hit the right notes for initial market
> creation, largely by observing what didn't work with Newton and
> General Magic and doing exactly the opposite, with some minimum of
> user input and feedback and a maximum of good intuition.
> Arguably VCs and investors are a kind of product alchemist; or
> perhaps more like patrons of alchemy. They have some of the same
> interests but different practices and goals; their focus is
> different: they vet product alchemists. One major area of focus for
> these guys is understanding market context, opportunity, execution
> ability, and --- particularly --- friction / barriers to adoption or
> success. The failure of this group of folks to come up with any
> repeatable process should be clear; in the end, it's all gut-calls
> and personalities, and the failure rates reflect that. Their major
> conceit: the barrier to entry. The inevitable conclusion:
> everything is impossible.
> And here we're mostly stuck. For three decades now we've been trying
> to build a science of product marketing out of alchemical parts and
> don't really have a unified discipline, yet. From the early days of
> quantitative usability studies at the (MIT) Media Lab back in the 80s
> to the abstract study technology adoption and innovation (i.e. Frank
> Bass --- whose work stretches back to 1969 but was little-recognized
> until the 90s) through the 90s (Clayton Christensen, and so on) to the
> object-focused alchemists of the late 90s and Oughts (Don Norman,
> Steve Krug, the Web 2.0 / 37 Signals crowd, etc.) we continue to get
> bits and pieces of the discipline but still lack any grand unified
> theory. And in fact, any such grand unified theory would likely be
> enormously (perhaps intractably) complicated. In addition to the
> above, it's also going to have to have some theory of actors and
> motivations, some kind of decision theory, some kind of theory of
> memes, and some well-formed construct that ties together the "outside
> context" parts of things, i.e. market context, path / history
> dependency, friction, predictions about the reactions of existing and
> non-yet-existing competitors and actors, etc. Makes e.g. forecasting
> in "mere" economics-at-scale look trivial.
> -- The Scientists ---
> There won't be any.
> -- The Post-Scientists / The Empiricists ---
> There won't be any science of product, no grand unified theory of
> product creation / innovation / marketing --- not because it's
> impossible (though it might be) but because we're going to leap right
> past that to something fuzzier, spookier, more massive, more
> inscrutable, more data-driven, colder and yet simultaneously "wetter"
> and more "biological" --- and more effective. Bigger, yet less
> substantial; ectoplasmic. Something much... Google-ier. Cf. "The
> Unreasonable Effectiveness of Data," "The Google Way of Science,"
> "Science Without Theory," etc.
> The basic idea here is: stop guessing, stop "modeling," stop
> extrapolating, stop forecasting, stop focus grouping, stop asking...
> and just measure and evolve. Measure early, measure often, measure
> always, measure everything, measure a *whole lot.* Let the numbers
> guide the decisions about what to build, how to build it, when to
> build it, *why* to build it. What color should that background be?
> Try them all and measure it. Should I build X, Y, or Z? Build them
> all, see which works best, discard the rest. Should that button be
> here or there? Let the (very large) masses decide.
> Google is the pioneer and the current (and foreseeable) master of
> this. Their scale coupled with their stigmergic processes and
> disorganization allow the exploration of very many possibilities in
> parallel, through a kind of evolutionary process. What works,
> propagates. What doesn't is culled, absorbed, refactored, and any
> good bits digested and reused. Everything is measured --- again and
> again and again. Nothing is stagnant; no solution assumed
> permanently optimal; mutation is a constant. Everything is
> continuously tested for fitness; that which is (currently) best is
> selected and used. New niches to invade and occupy are constantly
> sought. Meta-technology in action.
> They aren't alone in starting to do things this way, but they're
> certainly driving it forward as fast as possible.
> I have conflicted feelings about this. Everything reduces to the cold
> equations, the simple facts, the raw numbers, and the Really Big.
> Artistry is deprecated, intuition obsoleted, large scale made
> essential, and incremental bootstrapping from the tiny made difficult
> or impossible as the ability to conduct the process at small scale is
> limited. The world bifurcates; the distinction between craft
> products and tools will become ever clearer. I'm not sure that's a
> good thing; maybe it is. The tools should be more useful, the market
> overall should become more efficient at delivery of useful goods and
> services at the right times. And perhaps the disillusionment that
> goes along with such distinctions will eventually enable software and
> other technological *artisans* to conduct their trade for what is is,
> free from the idea that there's any valid measure other than the
> aesthetic / subjective / qualitative satisfaction of a limited
> audience. (Interesting tangents: what about the application of this
> sort of thing to e.g. investments? Data too limited today, within any
> given fund or even across funds and industries. But, maybe... Also
> worthy of consideration: impact of post-scientific product marketing
> on e.g. bazaar vs. cathedral models. Obviously, magic and alchemy
> occur in the cathedral. But post-science also has to, necessarily,
> occur in the cathedral today, because that's generally where the
> critical masses of data, computing resources, and so on reside. If
> only there were Turing Awards for product marketing: "Can
> Post-Scientific Product Marketing be Liberated from its Cathedral
> Style?" If so, how?)
> Nonetheless, that's where we're at.
> "Welcome to the brave new world of post-scientific product marketing.
> Your color is #C2D9FF. Enjoy it! We know you will."
> --- Beyond-the-Beyond ---
> For now, humans remain in the loop, the analysts and actors, human
> judgments and proactivity still essential to acts of invention, to the
> product-conceiving and product-making creative process. That won't
> always be the case. The next meta-level up the stack is automating
> away that part of the process, closing the loop entirely, humans
> merely as reactive agents and input signals, the population as a whole
> an experimental testbed on which the process operates. Quo bono? Us,
> hopefully. But not entirely certainly, and perhaps not forever...
> At the end of the line, we will find ourselves back to magic. Clarke
> was right, of course; "any sufficiently advanced technology..." When
> the technology in question is the meta-technology of technology
> creation, and when you've moved beyond theory to pure, inscrutable,
> dense, automatically-derived, data-driven mathematical models,
> predictions, and processes... the creative process is out of our
> hands, becomes recursive... the exclusive domain of focused (if
> global, hopefully friendly (*cough*)) optimizers, optimizing away,
> refining, perfecting, chiseling away at product-space oh-so-efficiently.
> And so we have a world of agents and daemons roaming invisibly through
> virtual planes that intersect the desert of the real at select
> touch-points, a whole bestiarum vocabulum of artificial, narrow beings
> so eager to please: barely-aware fabs and repraps and reality
> printers, vision-projectors and consensus hallucinator-facilitators,
> stock feed logisticians, meta-compilers, automatic user interface
> builders and mass customizers, usage statisticians and reality miners,
> attention heat-mappers and points-of-sale streamers, all manner of
> other localizers and trackers, transaction and other pattern
> recognizers, time-on-page and other attention-attenders,
> similarity-clusterers and preference-predictors, action-modality
> specialists and subculture classifiers and fashion advisors and
> memetic imagineers, cluestream-sniffers and cluehounds and sensor-node
> watchers, remembrance agents and interest detectors, gaze gazers and
> meta-suggestive sell suggesters, and so on... more attention and
> effort and urgent need-to-please lavished instantaneously yet
> continually on each individual than the sum of human attention and
> effort throughout history...
> A swarm of cold intelligences operating for and on us, navigating,
> searching silently and tirelessly across vast multi-dimensional
> fitness landscapes of shimmery, roiling, chaotic, noisy, raw, rich,
> pure data; mining it, sifting it, finding, reminding, refining,
> synthesizing... competing, negotiating, cooperating, mutating,
> replicating, mating, creating, minimizing the experimental error,
> "understanding" their human experimental subjects by reduction to pure
> math.. Climbing that hill... dancing some weird and intricate dance
> of creative destruction and destructive creation... Simultaneously
> the most personalized, customized, bespoke user experience for each
> individual, a cornucopia of products, a consumer's Elysium... a
> reality tailored perfectly and obsessively to each person, yet not
> lovingly; so impersonal, so cold, too... perfect. Flawless.
> Nothing objectionable, nothing out of place, no surprises - ever
> (except when you want one and don't know it. They do.)
> And the whole thing certainly begins to look like magic. Spooky,
> virtual, cold, alien magic.
> "Wonderful?" Certainly, if literally. Desirable? Arguable.*
> Inevitable? Probably.
> * Me? I guess I dig the idea. Count me in. But then, meatheads
> never impress me much anyway. ;-)
> FoRK mailing list
More information about the FoRK