Skip to content

AI Augmentation Symbiosis

Bertrand Russell laments at the end of his major work on human knowledge that “All human knowledge is uncertain, inexact and partial” [1].
This is however, precisely where human intellect comes from; the inexact and flexible connections are how humans operate in a fluid world where exact rules would be too cumbersome.
This points to a valuable disconnect between developers of knowledge work tools and academia, in how connections are thought about. Knowledge work tools tend towards the exact and neutral, such as defining ontologies, taxonomies, creating knowledge graphs and linking items directly, such as Doug Engelbart’s ‘high-resolution’ linking. Academia aims for rigour but rigour using the human brain’s fuzziness–this is why academics do not like high-resolution linking, they want the student or researcher to have fully digested the source material so that they can summarise in their own way, not ‘simply’ thread connections.
High-resolution linking does not show that the author invested thought and gained an understanding but it does help the reader (both human and synthetic) navigate the information so a ‘best of both worlds’ interaction should be investigated. This could be, for example, the author being able to point to specific ‘highlights’ or ‘examples’ in a citation, maybe by expanding what Reference information can hold.
It is also relevant to being AI development into this. AI development is currently focused on humans providing information in structured form for the AI system or ‘purely’ machine learning based where the AI is giving no knowledge which could taint their learning from the ground up. In order for AI to manifest and operate in the real world however, it needs to be able to synthesise information to degrees us humans can[2], which, as noted above, is exactly what humans are good at, so why not use the principles of how to improve AI to Augment Intellect in humans? How can we use the best of AI to augment the human brain in symbiosis(a) [3], not naive competition?
Much useful work is already being done by AI, including speech recognition, picture classification, document analysis and basic self-driving cars. More useful work by AI will be welcome as assistants to humans both to remove more ‘clerical’ work and as thinking partners helping solve vastly complex problems by ‘intelligently’ displaying relevant information to the human. I personally don’t see it as a positive thing to work to augment computers to the point of us humans becoming obsolete. We should not outsource our thinking, decision making and hence, our morality.
We need to drive technological information interaction to truly augment the human. We need to use the precision AI provides in concert with our wet minds. We need to provide feedback loops where the human is fully engaged. More like a modern video game than a cartoon where the robot ‘does the work’.

With laptop screens comfortably covering our foveal vision with high resolution information and the capability to do multiple trillion ML operations per second we can do this, and we can do it powerfully through the interactive symbols we call text. Let us augment our text to augment our thinking, let citations be simple and robust, let text describe itself for rich interactions of information connections and views. Let text be the gateway between human and machine since text has always been, and is inherently, the gateway between the fuzzy, wet–and creative, insightful–human brain the the exact, logical, grammar and rule following lines on a page.




It surprises me that researchers who work in this area either don’t have children or are not interested in the development of children since so much could potentially be learnt of how humans learn and grow. We are not born with blank slates, nor are we born without a body–we are very much embodied and our senses develop in tandem with our interactions. In simplified terms, we learn what vision tells us by phrasing about and correlating what we see and when our arms bash into something. We are organisms composed of organs. As we grow older we thrash about in different ways, such as our beautiful 3 ½ year old ‘baby’ boy Edgar, who now chooses a new persona every 15 mins, from Elsa to whales to T-Rex, where it is very important to him to get it right, including how many talons/fingers the t-Rex has. His mind is all over the place, organising, changing, learning, questioning and moving about in a world of soft relationships, not linear graphs. This softness, afforded by the chemical and electrical wiring in our brains, is a far cry from silicon. Let’s not present otherwise and let’s not let one be jealous of the other, let’s really use the power of silicon for its strengths, in symbols with our fuzzy, soft creative brains. I just wanted to leave that in here.



a) J. C. R. Licklider wrote in 1960:
“Man-computer symbiosis is an expected development in cooperative interaction between men and electronic computers. It will involve very close coupling between the human and the electronic members of the partnership. The main aims are 1) to let computers facilitate formulative thinking as they now facilitate the solution of formulated problems, and 2) to enable men and computers to cooperate in making decisions and controlling complex situations without inflexible dependence on predetermined programs. In the anticipated symbiotic partnership, men will set the goals, formulate the hypotheses, determine the criteria, and perform the evaluations. Computing machines will do the routinizable work that must be done to prepare the way for insights and decisions in technical and scientific thinking. Preliminary analyses indicate that the symbiotic partnership will perform intellectual operations much more effectively than man alone can perform them. Prerequisites for the achievement of the effective, cooperative association include developments in computer time sharing, in memory components, in memory organization, in programming languages, and in input and output equipment.”


[1] Russell, B. 1956. Human Knowldege, Its Scope and Limits. ISBN: UVA:X000337394. From:
[2] Marcus, G. & Davis, E. 2019. Rebooting AI. Vintage. ISBN: 9781524748265, 1524748269. From:
[3] Licklider, J., Man-Computer Symbiosis in IRE Transactions on Human Factors in Electronics. 1960.
From [Accessed 04 03 2021].

Augmented Text Software

Please note: When read in Reader, the citations in hard brackets can be clicked on to produce a pop-up with all the reference information and the Endnote references are superscript letters, not in brackets. They can also be clicked on to produce the Endnote.
Published inCorrespondanceDeep LiteracyDougDemo@50Dynamic ViewThoughtsWhy?

One Comment

  1. Nicholas Spies

    I have felt that a great weakness of hypertext/hyperlinks is that they have to be created manually. AI through Machine Learning could be put to work in the service of human understanding, even before it is capable synthesizing knowledge per se, by allowing machine-generated hyperlinks to directly link papers.

    At first this could be done by the relatively simple means of applying citations already included in the papers, and making hypertext links to the papers being referred to. This would enable a much faster review of current research than manually looking up citations. Of course the pool of papers would have to be made freely available on the internet to be entirely effective.

    Another level would be to give a large number of related papers to a Machine Learning program to let it discover relevant links between papers, based partially on explicit citations but also on implicit links to earlier work based on a textural analysis of text not explicitly cited.

    If this turned out to effective, it would allow readers not at first skilled in a field to get up to speed by referencing earlier research, already assumed by the authors of the research to be ‘common knowledge’ among their potential readers. This would potentially allow a wider range of people to understand current research by providing hyperlinks (an electronic ‘paper trail’) to earlier research.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.