When there is sufficient contact between two or more entities, over time, what started out as a Deterrence Based Trust relationship that was carefully framed by boundaries and edged with potential punishment now becomes a more equitable and better balanced Knowledge Based Trust (KBT) one.
Knowledge Based Trust takes time to develop (or requires a strong and clear reputation to be possessed by at least one of the parties) and it leads to a clearer calculation of trust assigned because motives, capabilities and reliability can be understood much better. What is more important is since this is a deeper level of trust, infractions do not necessarily break it, immediately because their context can be understood much better.
Most organizational relationships are at this level. The benefits of Knowledge Based Trust are that:
- Trust is not necessarily broken by inconsistent behavior at this level.
- This level relies on information rather than deterrence.
Clear, consistent and honest communication continues to be key here as it does, indeed, with any kind of trust-based relationship. While this is not rocket science the very fact that we are discussing the requirements and itemizing them suggests the difficulty of the task when it comes to implementing it at organizational and even commercial level.
Many large brands and businesses fail to move from the Deterrence Based model of trust they are comfortable with, because they can understand and quantify punitive measures and they can clearly see and measure boundaries and constraints. Their risk-assessment mechanisms are perfectly happy calculating all this and their risk-adverse nature falsely assures them that they have every base covered.
Knowledge Based Trust requires more than relationship building on autopilot based upon the assumption that neither party wants to lose the zero-sum game they’re playing. It requires an understanding of the person (or organization) you are dealing with and a deeper appreciation of what motivates their actions. It requires, as a matter of fact the moving of the relationship status between two people, a business and its customers, two rival organizations or even two nations from none to “In a Relationship” (for Facebook fans) knowing full well that what lies ahead, at some point, is going to become a little complicated.
One of the key advantages of Knowledge Based Trust is that in order to work it requires actual knowledge, of the facts that constantly inform the relationship. That makes it less subject to the fluctuations of emotional context and more likely to form a stable base for further development. Because facts are a lot easier to assess and verify than feelings or a sense of potential punishment Knowledge Based Trust can also work at a reputation level where trust is given without direct, prior knowledge or contact with the person or organisation.
We can choose to give our trust, as an example, to Amazon or eBay even if we have never used them before because so much is already known about them as a business brand and we are aware of the way they handle disputes and ensure everyone plays fair within their environment. We can choose to give our trust to the doctor checking our vital health signs based on what we know others have shared about him and what those we have talked to in the hospital have told us.
Knowledge Based Trust is an area companies like Google, IBM and Microsoft spend a lot of money researching because it can provide a basis for cognitive computing to operate on and trust to be calculated algorithmically both online and offline. To make matters even more complicated Knowledge Based Trust, as a term is used to refer not just to trust that flows between the boundaries of individuals and institutions, consumers and brands but also to trust that’s given amongst colleagues, friends and even work acquaintances who may share a common work environment or a larger work identity. We shall get back to it before this section is out but first we must look at one more application of Knowledge Based Trust.
Of even greater importance than Knowledge Based Trust flowing through the workplace is the kind of Knowledge Based Trust that Google and Microsoft believe can be calculated on the fly creating a handy and truthful evaluation of the authority of a website. While this may seem to be a little futuristic and probably not of any particular concern to us, as individuals or even consumers, nothing could be further from the truth.
We use the web on a daily basis to discover facts, find information, consumer news and carry out commercial transactions. In the offline world, whenever we engage in any similar activity we can use a plethora of signals some of which are overt and some of which are subliminal to determine the authority of any source of news, information or goods.
We know, for example, that the TV station that gives us our daily allowance of 30 minutes of news in the evening is an expensive entity to set up. Its buildings, equipment, news tracks and personnel make it difficult to fake and improbable to set up improperly because it will fail quickly and a lot of money will be lost. We also know that it is regulated by the government as it has had to purchase a license for the airwave frequencies it uses and it has competition in other, similar TV stations that also vie for our attention.
All of these observations form a thick layer of verification. While we may accept that the TV station may have a political slant, it is not going to lie outright to us, actively deceive us or try to trick us. And in the unlikely scenario that this very thing happens and our Knowledge Based Trust in the TV station in question is shattered there is an additional layer of Deterrent Based Trust behind it as we know there will be punishment for such behavior meted out by one of the real world recognized authorities (a government watchdog, the judiciary or the police) who themselves have a vested interest in maintaining the flow of trust throughout society so they can function.
The web is a decentralized, non-localized world where local authorities operate with difficulty. Its very set up inverts the real world dynamic making it easy to set up and maintain a presence cost-effectively. Within its environment it is easy to create a look that is every bit as convincing as any expensive, authority website and use it to spread misinformation, commit fraud or simply propagate lies.
This is where Knowledge Based Trust comes in. At the moment, websites are deemed to be trustworthy and authorities based upon the number and type of other websites that link to them, the kind of content they put out and the types of people who talk about the content, share it, comment on it and so on. The process is detailed and I wrote an entire book about the change called Google Semantic Search where I mentioned that Google’s index of the web is moving from “strings to things” understanding objects and people and places and names the same way you and I do.
This is important because a search engine could then also begin to assess credibility based on factual content the same way a person could. Well, better than a person would as a search engine will do it faster, tirelessly and will be able to call up at will the billions of objects and names and places and people (called entities) that it has in its index and cross-reference the interaction between them creating a complex, mini web of connections, each of which is ascribed a numerical value.
Fig 3.4 – The numerical value of the relational connections between entities on the web allows the calculation of relative importance each one has and the importance that it then gives to another.
At present this is a contentious issue the debate about which, paradoxically, is also based on trust or the lack of.
Because the web is becoming larger, denser and more complicated and the divide between the online and offline world is being eroded by the day it is worth actually understanding the issue behind Google’s KBT proposal. The paper, first published in February 2015 is titled: “Knowledge-Based Trust: Estimating the Trustworthiness of Web Sources” and it’s authored eight Googlers, amongst them Xin Luna Dong, a research scientist at Google. The paper begins innocuously enough:
“The quality of web sources has been traditionally evaluated using exogenous signals such as the hyperlink structure of the graph. We propose a new approach that relies on endogenous signals, namely, the correctness of factual information provided by the source. A source that has few false facts is considered to be trustworthy.”
But soon runs into hot water with its suggestion that: “The facts are automatically extracted from each source by information extraction methods commonly used to construct knowledge bases.”
Upon publication of the paper there was an article by Hal Hodson published in the New Scientist with the title, "Google wants to rank websites based on facts not links" and even searchengineland.com one of the most trustworthy websites for knowledge and facts on search engine optimization ran the headline: “Google Researchers Introduce System To Rank Web Pages On Facts, Not Links”.
Now, you might be forgiven for wondering what the fuss is about? Surely a suitably programmed search bot is capable of better ascertaining the truthfulness of facts and determining whether a website is trustworthy or not. But that’s exactly that, maybe it is, in most case but there must be quite a few where it won’t be. Facts and how they are determined is not always straightforward. Given infinite computing power and a suitably large volume of data the problem may be solved, but Google, just like any other business has to work with things in an ergonomic fashion that optimizes the computing power it uses for any given task so that it becomes sustainable.
Under that restriction errors creep in and because errors made by a bot are invisible both to us and to itself, our trust in a bot’s ability to understand it has erred and correct it is diminished. It is not beyond possibility to imagine a system that has run amok, errors accumulating on errors and providing wrong answers or branding as untrustworthy websites or entities that are anything but and because it’s automated, we have no recourse to setting the record straight.
In a smaller way the plot of the 1985 Terry Gilliam film, Brazil is about just such a nightmarish world where the system is simply inhuman and, as a result, inhumane. The perception that we may suddenly be at the mercy of an unwavering bot on a mission, directly affects our trust in the proposed results such a bot would deliver and, by the inevitable halo effect, then taints the trust we are willing to give to a Google branded search that uses such a system.
Two Microsoft researchers by the names of Moritz Y. Becker and Masoud Koleini, respectively working out of the universities of Cambridge and Birmingham, in the UK, thought about that exact vulnerability in our increasingly automated trust-based systems when they authored a paper in 2011 called Opacity Analysis in Trust Management Systems. Their contention is that where Opaque Trust Systems are concerned, an unauthorized intruder (or an error) need only get past the system’s defenses to then successfully present itself with the full authority of the system it has infiltrated. Their paper suggests some solutions to the problem but it is in the highlighting of something that really needs to be addressed in the very near future that it has its greatest value.
At the moment, Google has not implemented a Knowledge Based Trust approach to search and my guess is it won’t until it has solved both how to create transparency in the process from which the answers to search queries based on a fact-assessment system are based and how to allow us enough involvement so we can learn to trust the results. What is of interest to us is that it is lack of trust in the system due to its opacity that currently makes it fail and creates the wave of resistance it has encountered.
The Google researchers were aware of this perhaps. The introduction to their paper begins with a quote from the 17th century English theologian and logician, Isaac Watts, that says: “Learning to trust is one of life’s most difficult tasks.”
This is the point where we will look at Knowledge Based Trust within the environment of the organization but before we do, Google’s momentary failure with an automated system to mine and verify trustworthiness across the web gives us a formula for trust-formation whenever we launch a product or a service that entails cognitive computing of some kind. The kind of product or service that either emulates the working of the human mind or completely replaces it:
- Create transparency in the process
- Allow humans to be able to question the results
- Create a greater symmetry of power in the relationship between humans and the product or the process
Watts, of course, was right. Trusting has to be learnt. Particularly when the trust we require is one that has to take place in the workplace and requires colleagues to trust each other well enough to freely share knowledge.
IBM carried out a detailed study on Knowledge Based Trust that happens within organization. First, it established that organizations where the people who staff them share knowledge freely because they trust each other and the company they work for, do better in the marketplace, are generally more competitive but, at the same time, manage to be more open, transparent with s greater sense of fulfilment and generally happier employees.
We know by now what happens whenever we look at trust in depth and seek to analyze it. What seems relatively simple devolves into greater complexities, we find ourselves chasing wisps, phantasms, until we are left with what we began with: a quality everyone understand differently and uses similarly, that’s dependent on context and culture.
To avoid falling into that trap the IBM researchers created boundaries that allowed them to establish the criteria a knowledge seeker looked for when he considered whether to trust or not the source from which the knowledge came from. Publishing their findings in a paper called: “Trust and knowledge sharing: A critical combination” the IBM researchers discovered that knowledge based trust was further affected by two other types of trust: “…two specific types of trust that are instrumental in the knowledge-sharing process are benevolence-based trust and competence-based trust.”
They cited the example of where an employee may feel that a coworker knows the information that the employee needs (competence), but the coworker may not trust that he will be forthcoming at the time when the information is needed (benevolence). Conversely, the employee can be confident that there may be other people who are willing to assist the employee (benevolence), but these people might not possess the knowledge or skills required (competence).
The trustworthiness of a coworker then, according to the study, hinged on a combination of competence-based trust and benevolence-based trust that was calculated through five specific attributes or criteria:
- Common Language
- Common Vision
- Discretion
- Receptivity
- Strong Ties
Table 3.1 – What is notable and what should be evident by now is that despite the fact that we can pinpoint that trust is governed by attributes and can be broken down into types that have a specific effect the stumbling block, each time, or conversely the enabling mechanism is the human ability to communicate clearly, establish connections and form relationships.
Knowledge Based Trust, in the context that IBM studied it, is interpersonal trust that allows relationships between people belonging to the same group to form so they can function as intended.
Fig 3.5: Interpersonal Trust is formed from Knowledge Based Trust and it is a highly desirable requirement in organizations, the army, tribal groups, social groups, any kind of tightly-knit community.
The presence of Knowledge Based Trust in organizations gives rise to a high level of interpersonal trust amongst their members and creates cohesive units out of a loose bunch of people. Navy SEAL training at Fort Bragg in North Carolina in the US, puts recruits through such a rigorous training (week six is commonly known as Hell Week) precisely so that they learn to form, quickly the kind of interpersonal trust required for them to act and survive like a unit.
So, depending on your particular scenario:
- Knowledge Based Trust helps you provide an evidence-based approach to building and maintaining trust
- Organization leaders, CEOs, team managers and company owners can benefit from Knowledge Based Trust in their businesses but they must first provide the necessary conditions and boundaries that make it possible (team building, common language, clearly defined goals, a common purpose and an identity of the people within a business)
- Businesses of all types benefit by engendering Knowledge Based Trust and it helps them build their reputation but communication is key to their achieving that goal
Like everything we do in the world Trust is an evolving quality, as it accumulates it acquires its own weight of evidence and it morphs. It becomes less fragile, more empowering. It creates assurance where none would exist. Confidence where confidence is needed to make things happen. It becomes the lubricant upon which the gears of the world spin faster.
As we go from the initial moment of Zero Acquaintance where Deterrence Based Trust is required to safeguard us and create, at least, the conditions for a first contact to take place we create better, stronger relationships. These make us feel safer, smarter and stronger. They amplify our own power and safeguard us by removing whatever threat we might have feared existed beforehand.
Not every situation will follow the same trust development arc. Not everyone deserves to have our trust, no matter what. And even where we give our trust and establish relationships that are based on evidence, the Knowledge Based Trust we are prepared to give is different for each person and constantly changing for the same person, over time.
If that were not complex enough to make your head spin we will look at what happens when Knowledge Based Trust, over time, has a sufficient body of evidence to take the relationship we form with another entity (and I use entity here because it can equally well be another person, a brand, a company or –even- a device) for Identification Based Trust (IBT for short) to develop.
Excerpt from: The Tribe That Discovered Trust: How Trust is Created, Propagated, Lost and Regained in Commercial Interactions.