PKM Summit 2025 was a great opportunity to meet other PKM enthusiasts in a very friendly unconference setting. I only attended the friday of this 2 day conference but still was able to catch up with a number of contacts and met some more interesting people. What I especially like from this conference is that whilst we share an interest in capturing and linking knowledge, the backgrounds from the attendees are quite diverse. From engineers to solopreneurs, from corporate knowledge workers to keynote speakers.
Whilst there is a program and schedule, the format of the sessions is quite open. Many sessions are actually input from attendees, and this format definitely lowers the barrier to actually share some of your own knowledge and experiences. In comparison to last year, I do think the conference would benefit from a little bit of separation between the “keynote” talks and workshops / sessions, to ensure everyone experiences a baseline of session quality.
Other than that, I got what I was seeking from this conference: connection with people about the topic of PKM and new inspiration for methods and tools to deepen my own personal knowledge management experience.
Session: Jan lunge - Heaper.de
- problem
- big folder of files and images is slow
- media can also be folders that need annotation
- tags are not types
- solution
- media wrapped in metadata
- a heap of files with “blocks” that contain elaborate tagging
- pin tags to sidebar as sort of “folders”
- Tagging
- ai based tagging is great for searching, but not for exploring based on personal experiences
- tip: using mentions to add contacts to pictures where that person is nog actually in
Session: Marieke - human, are you still in the loop
- cognitive offloading
- the knowledge illusion
- asymmetry in what you think is in other people’s head
- taking a picture if a slide is a form of cognitive offloading
- critical thinking decreases when doing more cognitive offloading
- the knowledge illusion
- diverging
- tip: use LLM to generate different perspectives on a thought
- taking notes (marieke’s process)
- write perspective of author in own words
- own perspective add experience / from who did I get the idea
- expand on the idea
- couple to prior knowledge
- think about facilitating multiple type of thinkers
- immediate thinkers
- thorough / slow thinkers
- tip: try “Nature bathing”
Session: daniel berger ai code editors
- zed editor is favorite, uses Claude 3.7 for free
- windsurf flow
- finds relevant local files
- submit the full files to online LLM to add context to the original prompt
- sometimes get stuck when trying to answer a more detailed question
- Model Context Protocol allows data sources to expose API for integrating with LLM
- TODO investigate further Model Context Protocol
Session: Evelyn pacitti on Inspiration
- View on inspiration
- inspiration: coming from outside
- sparks: come from inside
- Gardening analog
- Cloud as inspiration: coming from the top, like a cloud
- Rain as filter: filtering as the rain that falls on the bottom
- Ground as sparks: bottom up, fertiliser
- Plants: stems, there are only a few plants going to grow
- Fruits: evergreen / valuable insights that grow from plants
- make a split bewteen inside and outside notes
- make a clear distinction between your own and others thinking
- dataview list all the sparks
- copy text out of dataviews to start new aggregated notes
- tip: matter as a reader alternative from readwise
- also able to listen to podcasts and highlight transcripts
tools to explore further
- raycast (spotlight on steroids)
- eagle (snip images to s3 / minio)
- heaper (self hostable pkm tool with strong image support)
- matter (alternative reader solution)
PKM Summit self hosting pkm talk submission
links
people met
talk schedule
Room 1: Dream
Kristina Mirchuk
AI-Powered Knowledge Capture: Automate, Organize & Retrievecredits
- picture by