"What is happening" is a description. "What should be happening" is a standard. The report is the gap between them. Most intelligence products blur the first and the last and skip the middle — which is why most intelligence products read as confident prose with unclear load-bearing claims.
Matchmaking Seeds:: #Self-oscilating-synthesizer, #feedback-machine, self-analysis
ShurIQ's DKR and Codiak continuous improvement teams, a,b, & c are inspired by Engelbart's decades of creating and iterating on these same ideas, which include a transformative mental model on the nature of how we work. Equally as important as v of the technological constraints and limitations that Engelbart and his team faced, was the challenging process of deconditiioning and unlearning a few seemingly axiomatic norms that are still entrenched within most businesses and brands today.
I have a controversial thesis that I'm still figuring out how to test within the scientific research program that we are running every night of the week, for ShurIQ. I'll sketch out the elevator pitch version of it here in this callout, so I can link to it and put a little publication pressure on myself to finish up the blog post this week.
![In the previous post in this series on KODIAK, continuous improvement communities, and other projects to emerge from Doug Engelbart's creative design firm, I provide a brief tour of the amazing, influential, and inspiring work of Engelbart, which served to increase the collective intelligence and group IQ of teams within any business, and was achieved using a number of Engelbart's inventions, software programs and protocols which where adjacent to hypertext, which morphed into HTML, to support knowledge workers in ways that steered the groups using them, towards better forms of communication, and using software as an augmentation layer for thinking, brainstorming, coming up with corporate strategy and encoding the various decisions, plans, schedules, resource management decisions, within systems that maintained a higher level view into the underlying intent and goals, which where expressed within the very same documents.
When I gush about Engelbart's work, as I discuss it's influence on ShurIQ, I'm sometimes asked, if it was so amazing and inspiring why isn't it more widely known, with the follow up, which is typically asked by the person in the room who is most skeptical of ShurIQ, who also becomes our grestest advocate and champion, after they go through their checklist of doubts and questions, did Engelbart's system deliver on its ambitious goals?
The answer is a resounding.. Sort of... Engelbart's designs, programs and inventions where some of the most widely adopted individual components within the 40 years where he was actively working in the field, in the form of the mouse, and in the most ubiquitous of his ideas, in the shape of Google Workspace.
Here's where the mandatory clip of The Mother of All Demos will have you asking if it's really good generative AI, as we see a preview of what was essentially Google Docs, with version control and active, live representation of two individual knowledge workers talking on their headsets, 40 miles apart, as they co edit a Google doc, on their respective work stations..in 1964!
Naming The Ideal State: A Nested Doll of Questions, with additional feedback loop analysis as each new answer is provided, as each additional piece of information contains valuable detail that, while typically important in a less granular view, providing additional plot dots, which do ultimately help to define trends and the direction of movement. Occasionally a single additional response from a user, wherein they are answering questions that have been specifically designed to help ShurIQ to understand the goal state of whatever projects, campaigns, strategies or scenarios that make up the user's intended use for the framework. ShurIQ uses a number of different algorithms of thought, MBA tactics for business intelligence, Socratic reasoning, and Sense Collective's 3x3 Inquiry Modelling framework, which helps to establish a matrix across 6 different anchor concepts, questions, or prompts, designed to help establish as much conditional logic tension and turbulence, within the reasoning and simulation models used to manage the frameworks various processe management protocols, which support ShurIQ's multi-agent framework as users team of agents take the completed responses from their Totem Persona Model, and transform them into Dashboard Modules which present the declared concepts, goals, risks, outcomes or milestones that have been extracted from the daily, weekly, monthly, quarterly and annual performance management framework.
ShurIQ's framework provides several different products and services which include:
Brand Intelligence Report with 30 Day Actionable Intelligence Recommendations, and are based on the scoring of the brand within their competitive set, a process that we demonstrate each week with the publication of ShurIQ's SPBI Weekly Report.
-SPBI Stack-Ranking reports are generated weekly for a number of business verticals including
- Vertical Drama Streaming Platforms - Sneaker Brands - K-Pop - Ai Agent Startups from last year's ycombinator Class - prediction markets - California Wines - Every client that gets an editorial Brand Intelligence report has the option to sign up for our monthly subscription to receive the weekly Stack Ranking report for their competitive set, which includes a realtime BI Dashboard with a number of add-on and a la carte offerings, that expand on the feedback loops that ShurIQ is so effective in identifying and designing strategies and insights that close the gaps which are surfaced through our deep research operations, which we run 24/7 as we develop our custom LLM that is designed to provide brsnda with even deeper integration with ShurIQ's multi-agent framework and our API which is designed to compliment our modular, purpose-fit plugin and app integrations, which csn be customized to fit clients precise needs as discovred during our intake and creative sessions held with clients as part of the different packages available.
Creating the weekly Stack Ranking reports involves a pipeline of different agent frameworks, schedules web search queries and targeted keyword tracking, event coverage with dedicated discovery and monitoring for trends, conditional logic where entities flagged within the brands initial Totem Protocol Persona design, trigger alerts and notify the responsible Ai agent that has been assigned to the brands account, in order to meet the requirements of the brands ongoing implementation strategy and the associated terms of that engagement can include the basic flat ongoing research operations included in the general subscription, or can be quite more elaborate, and part of a larger brief with Shur Creative Partners.
The half dozen vertical industries that we are currently focusing our initial launch programs on, maintain an information and news cadence that warrants running deep research operations every night, in order to catch any breaking news which relates to the scoring rubric and general conditions of the business vertical. As such, we have started generating nightly insights that are part of a computer science research experiment that uses Karpathy’s Autoresearch and extends the information management operations to include a 2, phase ontology design session that uses publically available information as well as any collection of source documents about the brand, including their brand Bible, recent pr packages, one sheets, white papers, and when dealing with highly specific industry Jargon, it is helpful to do the initial intake with as much recent social media content, industry coverage, interview with the public face of the company, senior executives. From this collection of artifacts we build out a general ontology that codifies the stakeholders, value flow dynamics, key marketing channels, sales channels, operations and company culture, to establish the brands Totem Persona.
Brands Persona has a dashboard where their profile is continuously being improved, refined, and broken up into distinct channels, each having the ability to create a distinct Persona name, distinct from the highest level persona, I.e
Acme Corp is the brand, and their primary persona is called Acme Corp.
They can also have a distinct persona for their Gen Z Marketing called Acme: Z And their Acme: FAQ channel, has its own set of communication rules, it's own grammar, ontology and rubric, as well as a file room for dumping whatever source material they want to use to train the FAQ agent, which becomes a digital twin persona, embodying the zeitgeist of how Acme answers questions.
The DKR Data Services Agency that we created today, with the Slack document ingestion pipeline, is connected to the Totem Persona system, in an intuitive way, with Slack being the obvious interface between these two modules in the framework.
The same way that we can continously train and create rules for managing all instances of the FAQ, and more broadly, establish a brand-driven system for how questions are handled, this same approach is used to establish the canonical source of truth and integrity for documents, decks, ad copy, and marketing strategy. Now, we have the factory computer system and punch clock, and fire monitoring systems for the factory.
You can think of what I demonstrated today, with the dashboard for viewing ingested document insights, as one of several Viewports into the ShurIQ database, which you can see encoding all this rich metadata, that annotates every piece of information, with whatever additional information we want encoded across these artifacts. This additional information isn't really meant to be read again, after it's encoded, at least not by humans, in this format. The dashboard does a crude job, in its earkiesg state, of showing how we can present all that data, and organize it into views, using the various InfraNodus type network graph visualization algorithms, which help us to see into the negative space, and force the graph into engineered shapes that represent the relationship between each data object, as inferred by all the other information.
Topographical clusters using the concept space, and the visualization of popularity, based off of metrics like downloads, number of followers, relationships, previously worked with x, went to school with y, spent more then $2000 in zip code xyz.
ShurIQ now is able to perform the same type of network analysis it does on a brands public conversation zbut instead, focusing on the environment in which it's behind the scenes operations are being planned, and decisions are being made.
This is a unique form of inside out reporting that we are beginning to do, and the resulting improvements in the underlying knowledge graph are fantastic, as ShurIQ begins to expand its deep understanding of the brand and it's business.
Additional insights:
ShurIQ is actually training and fine tuning it's business intelligence reasoning, as we provide it with lots of second order and conditional logic expressions, as we discuss and make explicit certain dynamics, relationships, tensions, friction, and constraints. As it refines its understanding, it is making its own edits to the knowledge graph, which I have been observing, recording and tracing back to its previous assumptions or recorded logic.
Providing clearer directions into what is actually important in the limited collection of documents we gave it do far, will go a long way towards improving its effectiveness.
The Last part of the dashboard layout is quite powerful, as a demonstration of a general primative that is a superpower: ShurIQ provides a confidence score along with every implied fact it presents, for lower confidence answers, it is also generating clarifying questions and giving us a space to provide answers which it will then use to update the knowledge graph.
We can do this across the entire foundation of esdh business verticals topographic space, as a means of bootstrapping the network, as we set sail for reaching 1 billion nodes of facts, recorded as Hyperdimensional triples, 6RDF node objects that produce reified annotations about annotations, combinations of Declarative Intent, joined with representations of ideal outcome, and cross-linked to the necessary resources, channels and stakeholders that are most likely to be put together in a manner that realizes the ideal state, at the best cost, or in the most profitable or benefits manner available.
Prediction markets, predicitive declarations, and integration with Shur Simulation Studio, with the addition of OKF and dotprompt wherein ShurIQ agents can bundle up OKF collections on specific concepts, hand them off to other ShurIQ agents, with the configured prompt needed to effectively use the tools and carry the level of context to realize the best outcome.
Simulated personas can produce artifacts that are then formatted into OKF collections and dotprompt prompts, which feedback higher integrity data, for improving the accuracy of the simulated environment mirroring real personas.
Naming the ideal state is the work. It's also the hardest part, and the part most analysts skip because it requires commitment. Once the standard is public, the description becomes evidence rather than opinion, and the gap becomes the product. The report writes itself when the rubric is real.
What this lets stakeholders do: read reports that make their epistemic commitments visible. Separate the description from the standard from the gap and disagree with precision at each layer.
What's still open: who gets to write the rubric, and how do rubrics stay accountable to the publics they're supposed to serve?