Demo@50 | Doug Engelbart | The Demo | What Is Still Missing | What We Are Building | The Collaboration | Who We Are



Gyuri Lajos



Who I am


My background was in physics, philosophy, and history and philosophy of science. I then became interested in the potential of computers to augment our problem solving capacities, arriving at the belief that without the augmentation afforded by computers philosophy is empty, and without computing philosophy is blind. 25 years ago I completed a PhD at Leeds University in “Language-Oriented Programming in Meta-Lisp”. After ten years I left academia, and have worked in IT as a software consultant ever since. For the last eight months I worked In Budapest as a Full Stack Javascript Lead Developer at Lufthansa Systems, and from October 2017 until February at Hubscience, a startup within the biomed knowledge management field.


At the heart of my research has been exploring what users need in order



My web or social media presence


I created a twitter account for MindDrive where I have piloted  the capability to turn Google Drive into a Personal Knowledge Graph. Also for TrailMarks itself:


On this archived page, from the Linked Up Challenge blog produced with the help of my co-author Andrew Benedek, we announced our plans to pivot Wikinizer, a “next generation personal knowledge management tool”, into MindGraph:


MindGraph is not just another app, it hosts an Augmentation Engine which powers the capability infrastructure needed to build DKR2, OHS etc. as I envision it.


I have 6 active private projects on GitLab Four are for my products, and two for the collaboration with Hubscience.


At Research Gate: I have copies of our published papers on WikiNizer together with my thesis.


What I am Doing


I prototyped a range of Personal Knowledge Management capabilities as a way of investigating the possibility of co-evolving a live bootstrappable kernel called Opidox, which supports full tinkerability, user extensibility, and personalizability on the Web. In its scope it is comparable to Dan Ingalls Lively Kernel


My kernel, instantiated at, provides a graph based Universal Data model. This model supports the design and generation of extendible Universal Hypermedia Formats and Applications, and it can also be used to incorporate entire connected neighbourhoods of nodes drawn from WikiData and other linked data sources.


I have developed prototypes which support web research, software design work, project planning, and tracking. Integration with the Interplanetary File System by Protocol Labs  promises a quick way to add team collaboration and full distributability.


How it Relates to Doug's Work


I learned about Doug’s work about six years ago. As Adam Cheyer, the author of Collaborama reminds us “One of Doug’s key ideas was bootstrapping, and that you should use the system you are building to build itself.” Collaborama carries this vision forward by applying it to software development itself. NLS was also built this way.


It is an overlooked fact however that NLS was bootstrapped to the point of offering a higher level of expressive power than the technology it incorporated.


By raising  expressive power bootstrapping saves the user engaged in articulating new capabilities from the intellectual unmanageability of unnecessary complications and repetitions. This is the key factor enabling us to pursue co-evolutionary change at a much reduced cost.


What aspect/features of the DKR I feel should be built


I strongly agree with the points that Adam Cheyer made in his answer to the same this question. I would like to add that we should not only try to build DKR, we should also take on the challenge of rekindling the Augmentation Research process itself.


The Augmentation Research that produced NLS was conducted on only one computer. But as Vint Cerf succinctly pointed out, what they built was a "one-computer world wide web".


Today, Augmentation Research should not be conducted on one computer, it should be distributed, globally collaborative, accessible, permanent, etc, from the outset.


I also firmly believe that we need to start with Personal Knowledge Management, and as indicated above, that it should be built with an inherently distributed technology which carries with it the prospect of instant collaboration. Fortunately IPFS, orbitdb, the whole nodejs / npm ecosystem, Protocol labs, matrix HashGraph supply us with the basic plumbing that can turn a “Personal DKR” based on MindGraph into a fully fledged collaborative and distributed DKR. I envisage place, we call “Conceptipedia” where discoverable content forming a public knowledge graph is made available with access granted via micropayments.


What infrastructures and components I think will be required and I am working on


Recently I found a Presentation by Doug Engelbart describing the capability infrastructure for OHS, DKR, Networked Improvement Communities doing the “C” task of continuous Augmentation Research.


The Capability infrastructure described in that presentation is virtually isomorphic to the capabilities I have bootstrapped over the years. Bootstrapping of this capability graph is in progress as I upgrade MindGraph's model to be compatible with the collaborative Common Knowledge Graph incorporating nodes from WikiData.


My call for action is: Let us build the DKR and the Open Hypertext System based on the MindGraph Augmentation Kernel, a new cosmology, that is fully distributed, and permanent, with micro payment, transclusion, all the things built in which Ted Nelson was right about all these years.


Remember that Doug’s vision, like Alan Kay’s, was not to simply to create an application, but to create an entire industry. Our current IT industry is built on the “Worse is Better” New Jersey Philosophy.

It is running us into the ground while software is eating the world. We need to reconstitute that industry by working towards the goal of regaining the lost Paradigm of Engelbart’s Augmentation Research.


How I Hope to Integrate with a DKR or Contribute


In practical terms, I propose getting the ball rolling by turning the DKR2 contributors questionnaire into a collaborative mindgraph. We can set it up as an improvement community so that everyone can elaborate their ideas and contribute to it by capturing relevant information via linking to entities drawn from the knowledge graph, from the collaborative mind graph, group annotations etc.


Integrate with Hypothesis, to start constructive discussion over the web. I have a prototype built called Trailmarks, that lets you capture and mark the trails you blaze across the Web as the user explores the frontiers of his knowledge. Its distinguishing feature is that it is serverless. I.e. if you wish it would work without ever leaving your home network.


On the other hand you can share entire research context comprising up to 5k nodes presently, in a single self contained HTML file that when opened turns into a live Trailmarks application.


This can also be hosted on the Inter Planetary File System...


To that mix I propose adding “Deep Annotation” which anchors annotations in The Knowledge Graph, and let’s communities of interests roll their own Knowledge Graphs as collaborative MindGraphs. Unlike WikiData or Google’s own graphs these will be built without the requirement of being notable. As such they will be of more value to communities and individuals with strong privacy guarantees, although still anchored and linked to each other through The “Common” Knowledge Graph, discoverable through Conceptipedia.


Over the past four months for HubScience I produced technology which enables a dedicated knowledge graph to be created based on biomed related data extracted from WikiData saved into MongoDb. The purpose of this was  to deepen their annotation process and remarks. I would like to make this technology “eventually open source”, and would be interested in working with the great team at eLife to add deep annotation to their freshly minted Hypothesis integration.


Revive the Project planner tracker piloted within TrailMarks. Get is started by making it collaborative using MongoDb, but in the meantime develop a migration path to IPFS.


I propose that we create a DKR2 group on, and shared with all the people in the group, so they can link to annotations they make or find within the group. I have the basic plumbing in place to turn an annotation group into a MindGraph compatible with TrailMarks. This integration could be accomplished with a month, and I would like to find community support for this effort.


With appropriate levels of support and collaboration all of the above could be completed by the end of July.


Beyond that, I would like to focus on bootstrapping and let the community do whatever they deem worthwhile. I would like to see integration with the matrix, riot and other things which may emerge.


My scenario/walkthrough


Instead of relying on Artificial Intelligence, use the Augmented Intelligence of the “MindGraph Matrix” (built with IPFS and As they pursue their interests, users will add nodes from their own MindGraph to the matrix, making them discoverable. As a consequence of discovering and connecting to things created by others within the MindGraph Matrix, the graph grows.


Nodes added to the graph are discoverable either by search or by following links within the MindGraph. The more people connect and use nodes created by other users, the more valuable the graph gets, and the more micropayments each user will receive. The networks of meanings emerging within the MindGraph Matrix, which is built for humans by humans, creates a bigger and deeper Knowledge Graph than can be built by any Artificial Intelligence. Collaboration not only creates better alternative Knowledge Graphs, it also help us to regain control of our digital future.


Let us recall that Doug’s dream was squashed in part by the by the promise of AI. Do not let the AI which currently exists squash human potential for the second time round 50 years down the line.


How I Prefer to Collaborate


Initially I would prefer to contribute remotely in a virtual team. I am exploring however the possibility of re-establishing contact with my alma mater, Leeds University. I am ready to relocate. I would be happy to act as a supervisor/external consultant for postgraduate students contributing to the DKR2 project or other relevant tasks. As for licensing, I am exploring the possibility of adopting an “eventual open source” model or “fair licensing”, as exemplified by GitLab.