Category Archives: complexity horizon

The why of how devops works


Business vs disruptive technology

Why do devops pushes go wrong?  Lets talk a bit about what devops is.  Devops is an attempt to merge the strengths of open source bazaar with the order and certainties of business hierarchy.  By clinging to the nomenclature of automation (as least as old as the water mill) as a complete stand in for a social phenomenon, organizations undermine their own efforts.  Businesses set the wrong organizational objectives in motion and then react to the inevitable failures by further detuning the successful components.

Why do companies seek out devops workflow techniques?  In the positive case they want to increase profit.  They are acting as futurists.  Improving their potential before it’s required.  In the worst case they are reacting directly to one or more complexity horizons.  Their social communication has been stymied or dwarfed compared to people communicating in functional code on the other side of the horizon.  They just can’t compete.

How people work

Many people have struggled with gift culture as the origin of open source.  Gift culture permeates open source, but is not it’s source.   Gift culture naturally expresses in any society where reputation currencies are in circulation and rewards are only expressed as probabilities.  In other words, gift cultures are the lubricant of free markets.  Once a person’s foundation of their hierarchy of needs is satisfied, people  work for improvements in reputation.  Accomplishments are the coinage of gift culture’s currency.

Bazaars are the most free, and have the most moving parts, so lots of lubricant is needed.  They are highly complex organic machines.  Adding components as they grow.  Their advantage is self healing anti-fragility.  Their disadvantage is huge pools of waste.  Individuals work to improve reputation, not for profit, because reputation currency offers better long term stability.  Most bazaar market machinery enjoys little or no economy of scale.  This is where devops can help a business reenter or influence a market dominated by the bazaar.  By merging the strengths of both.

Economy of scale of what?

Economy of scale is what automation brings to the devops equation, but it is not novel without nurturing gift culture.  What does a probability driven gift culture bring to the equation?  Imagination turned systemic.  Not for short term profit but for reputation and credibility.  Becoming known as a kind of problem solver becomes insurance (a mid and long term trust structure) against ostracization and obsolescence.  A guarantee of future (perceived) utility to society.

For a devops initiative to succeed you must also nurture gift culture.  To do this you must trust some appropriate objectives to with the people closest to them.  Then analyze their output for organization wide economies of scale.  This is typically done in codes (statistical or mechanistic) bypassing complexity horizons.  This is done to harness skills of the members of the society, as they try to build their personal portfolio of accomplishment.  With the grease available to them, workers can build the machinery the business needs, without having to express their solutions beyond code.

Traditionally business reserves access to objectives for the very top, and tactics are employed at the very bottom, with various strategies employed by middle management to glue them together.  This is eventually effective but requires translating all action to human non functional language.  This translation inefficiency is why businesses can’t keep up.  Most technical fields are in a race against the complexity horizon.  With clear objectives great strategy and tactics write themselves.  Why?  Code as a communication tool is fast.

Doing dumb things faster

This brings our first failure into focus.  A focus strictly on automation.  Automation is both positive and critical as it pushes forth the code created at the bottom of the hierarchy, embracing it in a way that creates economies of scale.  It is the end but not the means.  The means is worker access to corporate objectives and some freedom to implement them.  Look no further than Google’s policy of self directed projects to see how this works.  If you don’t nurture the gift culture both with recognition for good work and self directed opportunities to fulfill company objectives, you will miss great opportunities for new economies of scale.  Continuous integration really means technician access to objectives.

Ladders and snakes

This frames the secondary risk of failure.  Businesses are traditionally hierarchical because not all actors are trustworthy.   Most, but not all, employees seek to excel by improving their reputation. If you move objectives down the tree towards the bottom of the organization, it will present strictly selfish actors(sometimes psychopaths) with golden opportunities that they will take.  This justifies a metered scaling of moving objectives down.  This is where the failure lies.  Metering serves a purpose, to limit the destructive capacity of discovery of bad actors, but it is often used as an excuse to never push objectives to those expressing code.  A guaranteed failure for devops initiatives.  Potentially fatal to the whole business if it is already competing with a successful devops deployment.

As an important note psychopaths can be valuable intelligent members of a team, but they have special needs.  Rather than discard them it may be possible to create purely technical work roles with no direct reports and and no singular authority over data.

A well oiled machine

The misunderstood role of gift culture as the end and not a side effect may cause businesses to supplement perks for access to organizational objectives to far less effect.  Fundementally people don’t want to just feel like their reputation is good.  They authentically want to improve it.  Being able to both see AND influence most if not all company objectives is critical to identifying the potential economies of scale that the bazaar is often structurally unaware of.

The key to understand the difference between devops and mere automation is the complexity horizon.  Where functioning technical solutions are by far the most efficient way to express both better strategy and objectives, and retool objectives when necessary.  Technicians need access to objectives.  Then both the game theory centric corporate objectives and individuals long term reputation objectives can combine to create the well oiled machine.

About Civgene:  Understanding the currency of gift culture and the complexity horizon would not be possible without Civgene.  Civgene is a novel scientific theory which provides a framework to view all societies.  Please consider exploring it further to grow your understanding of human behavior, particularly in groups.

Open source and the complexity horizon


Open source really embodies three changes from typical hierarchical human social systems.  Gift culture, the right to fork, and perpetually increasing levels of complexity.  But these pieces are not all new, what changed to make open source happen?  What problem is it actually solving?

Gift culture is not new.  It is as old as currency.  Currency predated coinage as barter and skills and tasks.  So why did open source happen if not for gift culture?  What did change?  GPL.  A new kind of copyright license establishing the right to fork.   That’s how Linux, the trial of a hard right to fork based in law, succeeded.

The right to fork is not new.  Clearly established by the Christian reformation, and enabled by Gutenberg, the right to fork, until the 1980s was only established against ultimate authorities by war.  Civil rights in the freest countries acknowledged it and derived their rights from it, but did not explicitly establish as a basic rule of engagement and existence.  In politics threats of an ultimate fork were often sufficient to deter one.

What is new?  Complexity at modern levels is completely new.  Where is the complexity?  Not in tasks or problems to solve.  They are still simple to explain.  In communication.  In language.  What does language indicate?  Respect for stature and respect for others time.  Not always based in the currencies of accomplishment and skill, but as a product of many parallel societies.   A focus on the importance of social structure undermines ideas, there for innovation, and ultimately investment.   The social structure becomes impassible and no problems or tasks are solved.

Repairing social structure becomes a second level trap.  Meetings are held.  Seminars attended.  No, a fork is needed.   The problem needs to be solved in order to be assigned sufficient language to solve it.  The language to solve the problems have no parallel and therefore no linguistic identifiers for needed concepts.  The industry tries to solve this by pumping out new names and acronyms, but they are often the property of someone and useless for general progress.  This is a distraction.   Undeveloped ideas are slowed by the work needed to name them.  In the computerized, Internet connected world, the source code is language of progress.

The complexity horizon is reached when the task is so complex that less efficient top down problem solving can no longer function.  No amount of time spent can solve the problem from the top.  ‘Leader’ understanding doesn’t scale language fast enough.  The client can solve the problem better if administration doesn’t block him.  No right to fork means the client no longer invests.  Trust (predictability of future trends) is lost because their personal experience is impassible.  Future investment is diminished.

The perception that the ability to understand a problem and articulate it are always equal is a lie.  Therefore the complexity horizon occurs when comprehension of tasks outrun articulation of it.  False cooperation becomes apparent (bogus reciprocity) and destroys trust.  How can understanding outrun articulation?  The subconscious must participate in solving the toughest problems.  That is imagination.  Rationalization of conceptualization is being outstripped.  The metamind is doing the work but the rational mind and the mouth can’t keep up.  If a fork can be had, the solution can employ more minds at the task of articulation.  If it can’t the relationship between solver and the client grinds on failing to economize and destroying the trust needed for investment from both.

This would have been impossible to decipher without first exploring Civgene’s explanation of the metamind and it’s subconscious roles.  Based in fundamental behavioral contrasts between humanity and the animal kingdom, and the implications for economics.  Please explore those ideas at your leisure.

To help grasp this here are some practical applications of open source and roughly when their complexity horizons were reached.  Note that the open source alternatives begin to gain momentum at the complexity horizon but are not accepted as inevitable until some time later.

Linux: 1993-1995

The original, complete, experiment.  Operating systems are a software layer between varying hardware and the programs people are really trying to run.  Commercial operating systems were plagued with bugs and suffered from declining stability.  The cause was the non linear growth in variety of hardware a computer could be built with.  Communicating in code eliminated grafted societies and their cumbersome verbal language.

Bitcoin: 2008-2010

Currency looses it’s value to a client as quality of transactions become less visible.  The increasing non linear complexity of derivatives makes understanding any market impossible, ultimately damaging trade.  By solving they Byzantine generals problem the complexity of language is eliminated.  A small collaboration of solvers can write code to track and transfer currency for clients at a global scale in a transparent way.

Devops: 2013-2016

Internet applications are a way to handle reliability and scalability problems.  The non linear expansion of global cyberwarfare, and the non linear expansion of the internet of things (ultimately internet connected computers in all equipment) requires management of operating system functions at the network level or a systemic scale.  This seems to be the first multi-factor complexity horizon.  Devops holds another distinction as well.  It is a new system.  Not drawing on errors from past attempts to breach the horizon.

3d printers:  Soon, perhaps some breaches now.

An epic confluence of complicating factors defy description and add complexity for manufacturing in on demand customization, trade, natural resources, security, and in the race to the atomic scale.   3d printers are likely the first multi-factor complexity horizon with more than two vectors of complexity.  We have likely passed some of the factors already.