Collaborative opportunities

Possible collaborative projects

This is a list of possible projects for further developing DoView Planning and outcomes theory. Get in touch if you would like to collaborate on any of these, or if you like, just let us know if you are working on some of them independently. Also, get in touch if you want to explore other possibilities.

Further developing DoView Drawing Rules

• Further developing the existing DoView Drawing Rules. We already know that DoView strategy/outcomes diagrams created with our DoView Drawing Rules effectively capture strategy and can support better decision-making in many different situations. We want DoView diagrams to become a widely used rich visual language for working with strategy. One that captures all of the key information that decision-makers need immediately available when making high-stakes strategic decisions.

• Developing standard ways of representing the underlying evidence for ‘This-Then’ claims being made in DoView. This is a subpart of enhancing DoView diagrams as a visual language capturing any type of strategy.

A standard for storing and sharing DoView diagrams

• Developing a standard for storing and sharing DoViews. For instance, at the moment, our legacy app saves DoViews in XML format. An XML schema, or other format, could be developed that contains all of the key elements that would be included in a DoView. This would enable portability of DoView diagrams between different apps, platforms and systems.

• Version‑control system that snapshots DoView diagrams over time, enabling before/after comparisons and archiving for institutional memory.

Shareable Collections of DoViews

• Developing collections of DoViews that capture evidence-based best practice in different domains using the standard format of DoView diagrams. Doing this would speed up best practice sharing as others could ‘borrow’ best practice to include in their own DoViews.

DoView-creating AI prompts

• Creating prompts that can be used with other AI systems to create DoViews. We have shown how this can be done with our ChatGPT DoView Drawing Prompt. The same could be done for other popular AI systems.

Embedding unique features of our legacy outcomes app in other apps, platforms and systems

• Embedding DoView Planning and the unique features of our legacy app in other apps, platforms and systems. Anyone is free to include the unique features of our legacy outcomes app into other systems (with acknowledgment). We are happy to work with you on this if you want.

Using DoViews to abstract ‘This-Then’ claims from any knowledge product or domain

• Exploring the use of DoViews in quickly summarizing, abstracting and improving understanding of a wide range of knowledge products that are relevant to taking action in the world.

• Developing browser plug-ins that enable AI to immediately convert the content of any web page into a DoView, surfacing the ‘This-Then’ claims being made on that webpage or in an article.

• Comprehensively DoViewing current theory and knowledge in different scientific domains. Scientific domains can be regarded as a set of ‘This-Then’ claims. As such they can be represented in the form of a DoView diagram.

• Developing a platform for creating large-scale DoViews of a scientific domains such that researchers could contribute to it, identify portions of the DoView that their current research projects are focusing on, and amend it in the light of their research findings.

• Automatically grading the strength of each “This‑Then” link in a DoView using uploaded study metadata (effect sizes, quality scores), color‑coding confidence levels.

Real-time DoView generation

• Developing AI generated real-time DoViews of presentations such that, as someone is presenting, a real-time DoView could be being produced showing the ‘This-Then’ claims the presenter is making.

• Regarding the previous idea, having AI evaluate in real time the existing evidence for the ‘This-Then’ claims a presenter is making.

Further developing outcomes theory

• Further developing outcomes theory as a generalizable theory that can be used to conceptualize any outcomes system. An outcomes system is any system for identifying, prioritizing, aligning, measuring outcomes, implementing action, attributing change, setting accountabilities or assessing impact in any setting.

• Exploring the use of category theory or other mathematical approaches to mathematize outcomes theory. This would enable formal specification of its principles, ensure internal consistency of the principles within it and possibly identify new principles that logically follow from the current principles of outcomes theory.

DoViews in AI management, transparency, alignment and control

• Exploring the use of DoViews in AI transparency, alignment and AI control as a way of specifying and quickly communicating the outcomes, lower-level steps and risk management strategies any AI agent or system is using.

• Specifically, having AI systems and agents represent their interpretation of a user’s text prompt in the form of a DoView diagram. The user could then confirm that the AI system understands their full intention. They could also use it to tweak what they want to the AI system to do. An AI agent could then report back on the progress it is making on the DoView it initially developed.

• Developing methods by which other AI systems could audit whether an AI system is conforming to a DoView it has agreed to implement (as outlined in the point above).

• Specifically, using a watchdog AI agent to monitor a subject AI agent’s real-world action and map each of its real-world actions back onto the relevant box in its DoView diagram. Automatically traffic-lighting each box in near-real time, raising alerts if the subject AI agent’s behaviour drifts outside the approved pathway, and storing an audit trail.

DoViews in grantmaking

• Developing a platform for grant-making organizations to publish DoViews of the outcomes they are seeking and have grant applicants specify what they are trying to do by offering to develop subparts of the organization’s overall DoView.

DoViews in regulatory contexts

• In regulatory contexts, using DoViews to specify the regulation-related outcomes that companies should be seeking and the lower-level steps and risk management strategies they should be employing. Auditing agencies auditing organizations for regulatory compliance against such DoView diagrams.

New ways of communicating and collectively working with DoViews

• Storyboarding a DoView. Improving communication of what is believed should happen regarding a DoView, by storyboarding it and guiding people through its boxes with a brief multimedia overview of each step in the DoView.

• Experimenting with VR spaces for representing and collectively working with DoViews as user-friendly ‘shared thinking tools’.

Using DoViews as ‘strategy graphs’ to structure incoming and outgoing communications

• Automatically creating large-scale DoViews capturing all of the action an individual wants to take in the world in any area of their lives.

• Using DoViews created in the point above as the basis for organizing communication streams about any kind of project or action. Besides receiving your texts, emails, social media updates, and Slack messages as usual, you would also be able to see them mapped to the relevant sections of your DoView. This takes advantage of DoViews capturing the ‘strategy graph’. As new messages come in, they would be divided into segments associated with different boxes within your DoView. You would then be able to view all your communications within the context of the specific DoView section you are working on, and send replies directly from the relevant box as you decide on your next action.

Using DoViews for ‘What-If’ simulation

• Building tools for using DoViews for simulating different possible strategic pathways as you move through a DoView. Explore turning DoView diagrams into an executable models (Bayesian network, Markov chain or system-dynamics stock-and-flow). Letting planners run Monte-Carlo or What-If analyses, see probability distributions for downstream outcomes, and test which levers give the biggest impact before changing the real system. 

• Developing DoViews to specify branching chains of ‘This-Then’ strategic logic, allowing the system to ‘lock in’ certain sequences of ‘This-Then’ logic to be executed if particular What-If scenarios arise. This would accelerate decision-making in time-pressured strategic decision-making situations.