The internet has no trust protocol
Martti Malmi has excellently defined some of the problems with trust in the global web of information.
Fake news
Fake news and AI generated content are a massive problem as there usually is no quick way to fact check the content and source.
The DeepNude is a prime example of how AI generate content can have a severe impact on peoples daily life. Imagine having an AI generate a picture of you without your clothes on, shared within the community of your living. Not everyone may be aware of this is a fake, and therefore it can be very embarrassing.
Fake news is usually founded in a real source of information but is distorted to fit a narrative. With the never-ending flow of information available on the internet, fact-checking everything is near impossible for a single person to handle.
Even professionals like journalists that are paid to do their fact checking are usually incentive to produce a lot of content and therefore just relay information that seems to be correct without fact checking it first. There is simply no time to fact-check when you need to be the first one to break the news.
How to combat fake news?
There is no good centralized solution, sorry. Here is why…
All information relayed from a person is always presented in a subjective way. Therefore reporting becomes an interpretation of the events happing. There are always degrees of truthiness in the interpretation, but that is defined by the receiver of the information.
Of cause, it would be possible to record everything, to get a full picture of context around the event and present this as information, but you still have the interpretation of the receiver. Recording everything is not efficient and is very time-consuming, so we fall back to reporting and subjective interpretation from the reporter.
Big social platforms are trying hard to govern the information on their platforms, therefore they set up rules that are used to filter information. The problem is that the rules are objective and not subjective, and to comply with the laws of many counties in general, the rules usually ends up stricter than any one country laws.
So fact-checking information before relaying it is an interpretation of an intermediary that decides if the information is valid or not. This may not fit with the end receiver view of things.
Today most consumers of information have to rely on the good faith on the producer of information, as the time and effort to validate the information oneself is not worth the effort. This creates an environment where it is easy to produce information that heavily manipulated and very little fact-checked before it reaches the masses.
What is needed is a system that can immediately present the fact-checking of information in a context according to one’s subjective perspective. This requires a system that supports the possibility of subjective trust in a peer to peer way, resulting in a web of trust network. This way, the only information that is trusted within one’s network will be relayed or flagged. This is similar to whitelisting intermediaries that we trust.
Currently, many e-mail systems are based on a blacklisting principle where most spam e-mails are objectively flagged and removed; however, everyone who is not blocked already can send you an e-mail. The blacklisting principle has the downside that it is based on an objective opinion and therefore, even valid e-mails end up in the spam folder. Maintaining a filter yourself is very time-consuming and inefficient.
Imagine an e-mail filter that can leverage on your trusted network and only allow you to receive e-mails that are trusted by your network. When someone in your network whitelist or blacklist a new e-mail address, you will automatically also allow receiving or blocking from that e-mail address. This system enables you out delegate to a trusted third party the responsibility to help you filter information. The trusted third party can be family, friends, organizations, or automated systems, and the difference is that you decide who to trust.
Not only would it be your first-degree peers that you trust, but you leverage on your friends trusted friends, this can very quickly create larger networks helping filtering information.
Trust systems
Trust reputation systems in the digital world of today are mostly centralized and only within specific platforms. There are usually no sharing of trust and reputation between the platforms, and this makes it hard to keeping track of identities reputation across systems. Many platforms do even not have any form of reputation system in place, like many forums, news sites, and so on. Therefore keeping a personal record of each identity on many different contents providing platforms becomes a time consuming and near impossible task, as the identity of the same person may differ from platform to platform.
Many existing solutions only offer a specific kind of trust, like reputation and have no possibility for diversity in expressing the kind of trust a user may have of an identity or item.
Systems that focus on decentralized identity do not, in my opinion, solve the problem as the focus is on storing identity credentials locally and do not focus on why these identities credentials are trusted in the first place.
OpenPGP is a web of trust system that enables the user to share data with its peers in a secure way. However, it does not focus on a broader level where everything can be trusted within the context of a users web of trust.
Many systems that focus on identity and reputation, is implemented as an application and not as a protocol, this limits its application usability, as most companies will try to avoid being dependent on IP that they do not own.
There is simply no solution to my knowledge today that focus on solving the need for a web of trust system that can span across organizations and platforms, in a decentralized way.
Conclusion
A lot of the problems with the information on the Internet today is the lack of a filter that fits in the context of the observer. Besides problematic centralized solutions, there is very little a consumer of information can do to protect her/himself efficiently.
There is no way to maintain a proper trust network outside closed systems, that can cross organizations and platforms.
The Internet has no trust protocol.
