🌀 Technological Disruption Alchemy

🌀 Technological Disruption Alchemy

Almost a year ago I described the concept of γσ-machines (gnosio-machine). It was months before the rise of ChatGPT and the emergence of LLMs and other AI algorithms across productivity and collaboration tools. But since then, I thought I needed a practical example that illustrated it.

And here we go: I build a Miroverse template - Technological Disruption Alchemy that relies on the power of Miro AI to explore humanity's knowledge and generate ideas. It still requires manual work to connect the dots and operational modes, the AI model used by Miro is not the latest one (e.g., Bing Chat generates subjectively better results), etc. – but it gives that feeling of plenty of possibilities ahead.

Technological Disruption Alchemy

Technological Disruption Alchemy (TDA) is one of the Spectral Thinking tools developed to quickly catch up with technological trends and brainstorm new disruptive ideas on the frontier of modern possibilities.

TDA has two major stages:

  1. Context Retrospection and Prospection. You restore the past and future landscapes based on existing knowledge (history and foresight). Like what was on the hype? What did work and what didn't?
  2. The Alchemy Game. An exploration of possible combinations across three groups of ingredients:
    – The prospective technologies that you brainstormed during Stage 1
    – The niche technologies that are the foundation of your business today.
    – The magic technologies that do not exist today or are not scalable today.

The tool might be used for multiple occasions, but the main two are:

  • Consulting & Domain Analysis. Quick dive into a new domain: get oriented, understand key trends, and get some perspectives and insights on long-term domain evolution and potential strategies customers might explore.
  • Product & Tech Research Pivoting. Explore the options to pivot existing products or technology taking into account long-term trends and semi-random "magic" opportunities.

Context Retrospection and Prospection

The first stage is focused on your analytical mind: industry knowledge, technology potential judgment, and ability to forecast future stakes.

The template has three major parts fulfilled sequentially:
1. The timeline of ideas ranging from 1990s to 2035+. It is best to do prework here and fulfill the black stickers with technologies and critical products defining each of the epochs in your domain and its supersystem (e.g., mobile -> personal computing devices -> IT). Then at the session, you brainstorm with the team about what other technologies, projects, and turn points are worth mentioning. The discussion and context restoration begins with the past events and continues to the future. So you might need to consult with historical references or future foresight reports.

2. Judgement raws: you have to pick up key stickers to copy (or move) them to one of the rows - real disruptors if you believe that technology was or will be crucial for domain evolution, or over-hyped if everyone else believes but not you. You don't have to sort out all the stickers from the timeline – choose only the most important ones.

3. Top 3 prospective technologies. This is the final analytical exercise here – choose 3 stickers that will create the most dramatic effect in future: 2 from real disruptors and 1 from over-hyped technologies.

So we refreshed the technological context and landscape in mind and on the board, and now we are ready for the alchemy game.

Alchemy Game

The game has two parts, both rely on your creativity:

  1. Craft the TECHBOX. We need to prepare the ingredients for the recipes. There are three types of resources: PT - prospective technologies from the previous stage, NT - niche technologies that are foundational for your specific business (e.g., key tech in the product), and MT - the "magic" technologies - so advanced tech that it is indistinguishable from the magic. You need at least three of each.
  2. Complete the RECIPES. Next, you pick one idea from each resource group, put them into the formulas, and brainstorm what could be the result of combining all three or two. It could be a new product, new technology, new ecosystem, new concept, etc.

Sound relatively easy but easiness comes only with practice. So you might need some help. And this is where Miro AI comes to the rescue.

Consult Miro AI

Many of these tasks rely directly or indirectly on knowing the industry, technology development patterns, modern trends, etc. None of these is impossible to acquire, especially if you have relevant experts in the team or on the session. But what if not?

Then you might ask ChatGPT, or the Miro AI (I don't know which particular service and LLM the Miro team uses for this feature). To simplify that exercise I created templates with some prepopulated data. To use it there are two simple steps:

  1. Update the domain information to set some default context.
  2. Click on the final nodes in the mind map and use Miro AI icon to generate topics or ideas.

Finally, filter out irrelevant information and copy key ideas to the main board as new stickers.

🍄 As a side node, as with many other LLM-based tools you might need some prompt engineering to fine tune the model. Compare these two examples:

On the left I just declare what I'm looking for. On the right I added some prompt engineering kung-fu. I'm you could do better than me.

gnosio-machine prototype

Meanwhile, this template implemented over the Miro environment is still missing some critical features like structured domain context and editable raw representation (aka code), but it still creates the right flavor:

  1. The structured part of the template (white frames) represents some domain knowledge that might be fulfilled with expert knowledge. This structure guides me through generating new knowledge or point of view: individually or collectively.
  2. The generative part of the template (black frames) represents the connection to the external world knowledge machine (LLM) that might explore some facts or generate some ideas and bring them to me.

To make it all work as one I still need some manual work: ask Miro AI, copy-paste texts, fine-tune prompts, etc. But it already creates the right feeling at a distance, the anticipation of something big around the corner. Imagine the following:

  1. ✅ [Done] It is a template for Miro, FigJam, or any other collaborative environment that captures someone's experience on gathering and processing knowledge. So it is something repetitive, reusable, and scalable in that sense.
  2. The template has some variables that propagate through the board to make it specific.  E.g., domain, information sources, or timeline span. When the process is split across frames but some data must be carried over it might be useful even without the following steps.
  3. ☑️ [Partially implemented] The template has some visually structured parts that auto-label stickers once you put them in corresponding zones. It might be exported as data and edited raw (as structured text).
  4. ☑️ [Partially implemented] The template has some "generative" parts that rely on AI/LLM model with pre-tuned prompts. So once you fulfill the input you instantly get some output and might refine it, repeat generation, etc. Ideally, inputs are the "zones" in visual structure, a reminiscence of visual programming.
  5. The template has "autonomy": some of these tasks might be done and redone autonomously while the user is offline. So if LLM is updated or there is new external data you might get automatic updates and notifications. The template is the web, corporate, or humane knowledge crawler.

Technically, most of these ideas and features might be implemented with custom scripts or plugins tailored for specific boards even today. Unfortunately, it is not a scalable approach. Instead, all of these should be a part of the core engine. But it is another story.

In the future, we should expect such gnosio-machine bundles to be released by consulting agencies and strategy advisors with LLMs pre-tuned on domain-specific knowledge, embedded virtual AI-assistant with a hop to a human expert, and autonomous execution once configures with business notifications on new insights. That also creates new professional niches:

  • Knowledge/Industry Sherpa – a business analyst with domain expertise capable of prompt engineering, no/low coding, and data structuring (ontology) and visualization;
  • Gnosio Studio – a small agency releasing consulting gnosio-machines for specific domains: merges strategy consultants with data scientists and knowledge/insights visualization experts.

And that has mind-blowing potential to redesign the whole consulting sphere!


Miro/Miroverse: https://miro.com/miroverse/technological-disruption-alchemy/


The Technological Disruption Alchemy (TDA) tools, templates, and this tutorial were created by Constantin Kichinsky and distributed under a Creative Commons Attribution-ShareAlike 4.0 International license.

How to specify a license:

The "Technological Disruption Alchemy" tool by Human Spectrum Lab, Constantin Kichinsky, CC BY-SA 4.0