Dark Light


Written by Anthonio Pettit
Creator of Structured Speculation

Structured Speculation is a project I developed through participation in the Landmark Team Management and Leadership Program.

The past several years have exposed fundamental breakdowns in workability between factions of people over superficially opposing viewpoints. As a result it’s become much harder to see a path towards common ground.

I gave a lot of thought to discovering what it would for people with disagreements to still have a way to collaborate around the outcomes that are mutually desirable.

If two people can agree upon some goal in the future,
it’s in their interest to determine what it would take to make that happen!

What would it take to put aside their differences, maintain civility and work towards a successful outcome?

This principle lead to the creation of Structured Speculation.

Through this process two or more parties with different agendas can interact around understandable solutions for tomorrow which are sensible today—by individually contemplating how something might work, each group silos themself away from others. They focus on all the reasons things couldn’t work out based on what is known from the past.

A commonality across concepts, that satisfies both groups when created together produces new ideas. Even if these ideas originated on separate individual thought paths. They cross over at one destination!

Structured Speculation started out as a new project management methodology based on creating declarative futures. I had worked out a lot of the nuts and bolts of the approach, but I knew I’d need some sort of AI backend to make sense of peoples’ conceptual goals. As far as I knew, I was just looking for some kind of AI that could help to classify and categorize various projects.

Meanwhile, I stumbled upon GPT-3 quite accidentally while looking to solve a different kind of challenge at work; attempting to automate the categorization and ranking of the worst of the worst terms on the internet. That was the initial use-case I believe I used when I signed up for the academic license. It can definitely handle those kinds of classification tasks with a relative ease. You give it a few examples of words you’ve classified a certain way, then give it a bunch more examples it hasn’t seen and ask it to classify them the same way.

Soon I started putting it through its paces. The whole premise of Structured Speculation was to be able to ask: “What would it take for XYZ to exist?” With GPT-3, I was able to ask “What else would I need to design a project management methodology like this?” GPT-3 not only helped me author Structured Speculation; it simultaneously produced the algorithm that it uses to generate the Structured Solutions I’ve been sharing with you. I gave GPT-3 a goal, in this case the rules of a project management structure, and it’s able to make logical sense of the rules and apply what it’s learned from humanity’s greatest hits. The end result is astounding, often weird and wild inferences about whatever it’s being asked to complete. There are so many implications about what this can and will be used for.

Structured Speculation is a creative problem solving methodology. Structured Speculation helps identify the necessary steps to complete a goal, and identifies people or organizations that need to do these things.

One implementation of Structured Speculation is as an engineering framework for inclusive creativity and its aim is to build bridges across diverse perspectives, so knowledge can be made explicit rather than implicit.

A main objective in Structured Speculation (SSPEC) is building collaborative solutions rather than competitive ones. SSPEC provides a way reduce bias reactivity and create environments where participants have more opportunities for understanding each other’s point of view by navigating natural resistance that arises when challenged with new information or innovative ideas

At the core of structured speculation lays embedded principles geared towards addressing socio-cultural factors which work together to promote social justice and robustness of this process .

Structured Speculation is an iterative process where participants need to trust each other and provide input collaboratively. This creativity method not only has a higher purpose but also works as a revitalizing exercise which boosts one’s creative skills and can lead to generating more original ideas .

Structured Speculation helps develop the following competences:

  • Social intelligence, empathy, open-mindedness and awareness of different perspectives;
  • Creativity, self-awareness and curiosity through divergent thinking;
  • Collaboration skills: listening openly, articulating ideas clearly, sharing ownership of ideas.

The idea for Structured Speculation came to me after witnessing the utter breakdown in communication among factions of society over the past five years or so.

Logically, if two people want to see the same result in the world, there ought to be a way for them to work together to achieve it, regardless of whatever differences might stand between them.

If only there were some way for people to unite around the things they do want in the world, rather than bickering about all the things they don’t.

Structured Speculation provides a way for groups and organizations to do just that;

A Structured Solution builds pathways collaboratively through a three step process: identifying an optimized outcome, breaking it down into measurable criteria called conditions of satisfaction (CoS), and lastly assembling resources and pathways which allow these CoS to be met without harm.

Structured Solutions harness a variety of ideas from diverse perspectives—including knowledge learned tacitly via observations–to tackle problems holistically while circumventing common cognitive biases such as omission bias or looking for single causal factors when multiple factors may have caused an outcome.

Artificially Co-Authored

From inception, Structured Speculation was likely to need to utilize some type of artificial intelligence to help determine and generate the classification of projects submitted to the system.

What I didn’t expect is how ready and able AI already is to take on this type of challenge

The Generative Pre-trained Transformer 3 (GPT-3) from Open AI became my confidant, research assistant, and writing partner.

GPT-3 is an advanced deep learning language model which is capable of comprehending and generating original text.

The process was not unlike co-creating an art project. After forming a rough outline of each section, I fed it to GPT-3 and asked for recommendations and refinement. Not every suggestion is useful or relevant due to the stochastic nature of its output. However the good suggestions are often astounding, and with a little curation, highlight the power of computing now available to humanity.

The Doctrine for Structured Speculation is itself a proof-of-concept for the methodology it describes, and is a demonstration of a Structured Solution.
The resulting Structured Speculation algorithm takes all of the methodology into account and applies it to new inputs. As GPT-3 is pre-trained, it currently cannot access real-time data from networked sources.

However once the proposed Loci framework is integrated, and safety protocols are hard-coded, the platform to will allow Structured Speculation to tap into the collective genius of humanity, and generate solutions in a way never before possible.

The future of what could be – is ready to create now.

Structured Speculation ©2021 The Foundation for Structured Speculation
Conceived and authored by Anthonio Pettit

Built on the distinctions of the Technology for Accomplishment™ from Landmark Worldwide Team Management and Leadership Program and Communication Courses, all rights reserved.

is powered by GPT-3 from OpenAI

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

The work on this website and the term “Structured Speculation” is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.