My vision for Golden

I believe that Golden is taking a crucial first step towards the far vision of an ultimate knowledge network of everything that almost everyone will need to rely on everyday in order to get any kind of work done. Right now, Golden is focusing on canonical knowledge - hardly debatable facts with a near absolute probability of being true. However, that is just a small part of all human knowledge. Beyond that is the realm of more debatable propositions such as “Elon Musk’s acquisition is beneficial to Twitter,” or “Ukraine-Russia conflict leads to nuclear war,” that are neither 0% nor 100% true, and would require not only a vast amount of factual data but also insights from qualified people to determine their probabilities.

While Golden’s entities are flatly linked to each other with predicates, the units of probabilistic knowledge can be better presented hierarchically where each unit leads to/proves/supports others. The objective probability of each high-level less-certain unit in the hierarchical knowledge graph can be determined based on those of the lower-level more-certain units that support it, which come from 3 sources: qualified people, canonical knowledge, and big data.

Firstly, knowledge provided/proved by people with related credentials has a high chance of being accurate. Golden needs to build a parallel graph of human credentials, where each knowledge provider can submit their own credentials (certificates, videos, Quora answers, etc.), and attest to others’. Credentials would be categorized into the taxonomy and quantified between 0 and 1. A board of experts in different fields need to join Golden as initial high-credentials users who will attest to the credentials of other users. Secondly, the canonical knowledge graph is being built by Golden, where validation doesn’t take experts and can be performed by anyone with common sense. Thirdly, big data refers to data from all possible sources such as IoT, transactional & operational systems, CCTV & satellites, that extract information all the way down to when who went where and did what. One day such data will be commercially available on a marketplace, where traders can use data credits to trade and Golden tokens to rate data providers.

The endgame for Golden would look like: a big data marketplace, a canonical knowledge graph, and a human credentials graph feed data for users to build up insights layer by layer in the hierarchical, probabilistic knowledge graph, which are configured as an oracle network streaming actionable data to blockchain protocols. For example, you can execute planned trades when the probability of “Ukraine-Russia conflict leads to nuclear war” reaches a certain threshold. On top of that would be a layer to support decision-making, and allow decision makers to sync decisions/outcomes back to its data sources. This vision is closer to “Web3 Palantir” rather than “Web3 Wikipedia.”

At the core of that system is the hierarchical knowledge graph which holds the power to accelerate the expansion of the canonical knowledge graph itself. While Golden attempts to financially incentivize people to build the knowledge graph, Wikipedia has been tirelessly built by countless contributors without any financial incentive for two decades. The fact is people are driven by more than just money - cause, ideology, sense of purpose. If we can tap into all motivations at once, the knowledge graphs can expand exponentially faster, as people would deploy all resources (knowledge, time, energy) to prove them right and others wrong. Instead of randomly grabbing data and putting them in place, a hierarchical knowledge graph would allow users to prioritize the most important and urgent things they want to prove/disprove (e.g. economic or political hypotheses that affect them the most) and try their best to gather data/insights that matter the most, which enables interesting product offerings (campaign, comparison, pitch, election, investigation, accreditation, privacy, auction, etc.).

Personally, I used to spend thousands of hours envisioning such a product/business. Back in 2011, I was writing a science fiction and became a heavy user of Wikipedia, and then Quora. There I stumbled upon concepts like Theory of Everything and Technological Singularity, which are final destinations for all knowledge-seeking journeys. Then I became obsessed with mapping all human knowledge in a single frame of reference, along novel dimensions which serve as guideline for everyone to seek wisdom. In 2015 I attempted to design a product and business around that with the 2016 U.S. presidential election as the first use case, and called it PropMap (where each unit of knowledge props up one another). However, I found it hard to find like-minded partners, so in 2017 I pivoted and renamed it as DropDeck, which narrowed the scope down to a single use case - to build a reputation graph of startups and related people (founders, investors, employees, etc.) to help identify the most potential investments.

To illustrate the vision for Golden laid out above, following are past documents and videos made before 2016 when I didn’t have any product development or web3 experiences: