[[Seymour Papert|Papert]] introduces the idea of the knowledge machine in "The children's machine: Rethinking school in the age of the computer" [[@papertChildrensMachineRethinking1993|(1993)]], through the story of a 4 year old, Jennifer, who questioned how giraffes sleep. After discussion, Papert was able to throw himself into an exploration of giraffes but opportunity was not available for Jennifer who had not yet learned to read or conduct search at that age. After which, he goes onto pose that there is no technical obstacle to create the knowledge machine - a machine that put 'the power to know what others know into Jennifer's hands'. Papert noted that what separates ebooks and interactive videos from a true knowledge machine, is the effort needed to bring the together the knowledge - something which now seems to have been overcome through search engines and [[Large language model|LLMs]] (as of 23/10/2025). He hypothesised that such a system would empower Jennifer to explore the world, through speech, touch, or gestures, whilst steering the knowledge machine to areas of her interest (not necessitating even the use of words). This would allow her to navigate through a knowledge space much broader than any printed encyclopaedia. That is, a context that facilitates even a child to be an [[Autodidacticism|autodidactic]] learner led by their [[Learn drive|learn drive]]. The knowledge machine in some sense aims to go beyond words, not letting ourselves be limited by the modality of text for learning, and embracing [[Multimodal learning|multimodal]] representations such as interactive simulations, virtual worlds and more. Though for Papert the knowledge machine was not only a way for disseminating knowledge (akin to prior [[Instructionism|instructionism]]), but provides a [[Constructionism|constructionist]] environment that provided materials, artefacts, and tools that allowed learners to construct their own understanding through creative acts that lead to a tangible, public artefact. I'm excited about how the innovations with [[Generative AI|generative AI]] and [[Virtual reality|virtual reality]] devices could play out in better realising this vision, given that so much of the realisation of technology ideologies can come down to the smoothness of [[Human-computer interaction|human-computer interaction]]. As of writing (23/10/2025), LLMs have become quite powerful and now a lot of the innovation is happening around generating other modalities, and virtual reality devices are becoming better with the Apple Vision pro and Galaxy XR bringing strong immersive experiences and smooth interaction design with gaze. However, the weight needs to be reduced, local computing power needs to be stronger and cost needs to go down before adoption will hit - currently it is still too cumbersome for the mainstream. ## Memex The knowledge machine is a type of [[Memex|memex]] but with a value of learning and creation, over simply task performance or construction of knowledge for scientific purposes. The ideas here start blending together, between [[Memex|memex]]'s like the [[Knowledge machine|knowledge machine]], the [[World Wide Web|web]], [[Personal knowledge management system|personal knowledge management systems]], [[Autodidacticism|autodidacticism]], [[HyperCard]], [[Wiki/cards/Incremental reading|incremental reading]], and more. Hopefully one day someone will organise the qualities of many memex's and how they support learning (as of 23/10/2025 I haven't seen, please email me if you know of any). ## Criticisms The knowledge machine was likely a founding motive behind the [[One laptop per child|one laptop per child (OLPC)]] project led by [[Nicholas Negroponte]] that was built on Papert's constructionist ideas and disseminated millions of laptops around the world. Though OLPC acts as a case study to the flaws of the knowledge machine's ideals, with Ames ethnography pointing out the nostalgic design that is built upon the assuming a world of 'technically precocious boys' where many of the [[Autodidacticism|autodidactic]] assumptions did not play out [[@amesCharismaMachineLife2019|(2019)]]. It seems that technical research into EdTech has fallen away from supporting the autodidactic learning (as of 23/10/2025), to simply supporting formal education with slightly better feedback or the like. I believe that the primary flaw of the knowledge machine idea is the underestimation of the need to provide [[Self regulation|self regulatory]] support! In situations without a teacher or parent to guide and support the child, it requires not only gathering of knowledge and providing materials for construction, but to provide structure and regulatory support like is seen in [[Formal education|formal education]]. Even current [[Personal knowledge management system|personal knowledge management systems]] like [[Obsidian]] or [[Anki]] are primarily used by motivated university students who have to go through many iterations of experimentation to learn how to make use of it for themselves. Even then, they often use it poorly! We cannot expect students to self-study the literature on the [[Learning sciences|learning sciences]] to better aid their learning process - rather the [[Memex|memex]] should compensate for the lack of human support, without the fatal over-engineering and over-scripting that makes the technology lack flexibility (like [[Intelligent tutoring system|ITSs]]). Such a balance is hard to achieve. ## Supermemo knowledge machine Those who know me, know that I love the ideas of [[Piotr Woźniak]]. Whilst Papert's notion of the knowledge machine was focused around fetching information and constructing artefacts with a hope to internalise relevant knowledge, Piotr nails the foundations and goes directly to supporting human memory and cognition. Taken from [Piotr's doctoral dissertation](https://web.archive.org/web/20160305172704/https://www.supermemo.com/english/ol/knowmachine.htm). > ... self-instruction systems based on repetition spacing form another crucial extension to the Papert’s Knowledge Machine. After all, the progress of sciences, technological innovation and social change all originate in the human brain. The outcome of information processing in the brain depends on its factual and inferential information contents. These, in a vast proportion, come from what we learn in the course of our lives. That’s why the conscious choice of the learned material with the full rational management of the remembered knowledge form such a decisive component. As I showed earlier, repetition spacing is the only way towards this conscious choice and rational management. That is, the knowledge machine should also support the conscious decision to internalise knowledge, see [[Wiki/cards/Incremental reading|incremental reading]]. # References Ames, M. G. (2019). _The charisma machine: The life, death, and legacy of one laptop per child_. Mit Press. Papert, S. (1993). _The children’s machine: Rethinking school in the age of the computer_. Basic Books, Inc. [https://dl.acm.org/doi/abs/10.5555/139395](https://dl.acm.org/doi/abs/10.5555/139395)