Starlight will be building and introducing a new feature on the information it sources from news and other text based documents. We will be deploying this feature for natural disaster prevention operations with an initial sector use in wildfires. We will start with reputation based veracity confirmation of text documents in the news domain, but quickly scale up efforts to begin what will become cross-modal geospatial data confirmation of a distributed platform.
How does it work?
We have a small internal service for education called WritersBloc which allows teams to program one or more personalized-information bots to search Twitter on their behalf to collect educational documents and articles on behalf of the team. Each bot has programming to explore three topics to search and gather news articles and documents that have been shared and commented on by a well-regarded member of that subject matter domain, which we call Emblems. Today, each Emblem is given a simple reputation score based on our evaluation of their skills in the subject matter domain, and each Article is also given a veracity score on the basis of our perceived truthfulness of the content of the article. Each Publisher is determined to have a simple Truthiness score on the basis of their articles and the reputation of those who share them.
Emblems may appeal our determination of their reputation in various skills domains and we anticipate creating a skills pedagogy for Emblems and non-affiliated individuals to demonstrate their skills performance in a given area.
Publishers may receive a clear and transparent articulation of the equations we use to determine their overall Truthiness and a list of annotations made to each Article that explain our concerns over the truthfulness of any one article.
These are unrealized opportunities for massive transformations in education around the world. We imagine a service driving an app to deliver any content on any set of topics to any child around the world with age-appropriate, clean, annotated, and veracity-checked learning materials, and a framework for any child to demonstrate their early and evolving skills domains across domains.
The framework described above may be an early brainchild of the research division at Starlight, but we hope that by developing high quality, immediate resources for natural disaster first responders and their operators, we can develop a mastery of how to timely target clear and actionable geospatial data and text effectively, truthfully, transparently, and in a distributed platform.
We know this is an ambitious vision, so why are we taking the leader position?
Climate change and the social welfare of Americans and children around the world. We simply believe we need every child rowing their strongest direction.
We live in a world of ravaging climate confusion, and a tsunami of fake and false news arriving. We live in a land of skyrocketing mental health challenges brought on by social media, and increasingly challenging class inequity. We live at a time when two billion new children are about to be born, and one in five children live in poverty at home. Were this only a movie, we’d hear a speech now about releasing the hero in all of us. We’d want to be shining lights.
We believe the idea of a universal, distributed and consistently evolving library as the capital of knowledge and learning is an old and effective one.
There appears to be no more powerful force in the world for transformation of poverty into pathfinding, confusion into actionability, injustice into resolve, and strife into accord that individuals empowered to act on their knowledge.
We hope WritersBloc is the first step to promote this access for all the world.
Get to know us. Starlight.
David Bernat is founder and president/CEO of Starlight, a company dedicated to innovating intersectional technology with satellites and AI. Dr. Bernat is a Space Scientist and former member of the Goldman Sachs strategies desk.