# A Few Observations From... ![rw-book-cover](https://pbs.twimg.com/profile_images/1582166936607072256/owiQcbhQ.jpg) ## Metadata - Author: [[@RobertHaisfield on Twitter]] - Full Title: A Few Observations From... - Category: #tweets - URL: https://twitter.com/RobertHaisfield/status/1492656301939191810 ## Highlights - A few observations from this piece from @danshipper about why he isn't using @RoamResearch anymore: 1. If all links are intentional, then reviewing them is less likely to surprise you 2. Many people review notes so infrequently as to make upfront structuring effort not worth it ![](https://pbs.twimg.com/media/FLb3OuyVcAAKQBj.png) ![](https://pbs.twimg.com/media/FLb3q7SVEAEUQ77.png) ([View Tweet](https://twitter.com/RobertHaisfield/status/1492656301939191810)) - 3. It is valuable to remove the friction of "where do I put this?" as that can get in the way of writing the note (people like @EleanorKonik who have a well architected system and a sense of "mise en place" are the exception) https://t.co/muQ07Qelr6 ([View Tweet](https://twitter.com/RobertHaisfield/status/1492656303478489091)) - 4. Strong search can remove the friction of placement. Roam's design seems to want users to create new pages for terms & filter through unlinked references to find info instead of a more traditional search. This is good for exploratory search, inefficient for focused search. ![](https://pbs.twimg.com/media/FLb4axnVIAAi8XV.jpg) ([View Tweet](https://twitter.com/RobertHaisfield/status/1492656306565500930)) - 5. If people rarely *intentionally* review notes, upfront structure is not worth it. Additionally, it's difficult to predict what you will use that structure for in the future. Ideally, your PKM should be able to generate structure in hindsight from as little info as possible ![](https://pbs.twimg.com/media/FLb5gfcUUAIJDoT.png) ([View Tweet](https://twitter.com/RobertHaisfield/status/1492656309258231813)) - 6. The best tools for thought of the future will work like a programming language that makes it trivial to refactor information in hindsight (since you can't predict all possible ways you will use it) with "AI" that writes "code" for you. https://t.co/LU99LkYU6I ![](https://pbs.twimg.com/media/FLb7wfhVkAAmpEz.jpg) ([View Tweet](https://twitter.com/RobertHaisfield/status/1492657139365519361)) - Link to the original blog because I forgot in the original tweet lol https://t.co/fV8ykIN3SG ([View Tweet](https://twitter.com/RobertHaisfield/status/1492658135617531906)) - With everything I write into my notes, there's no reason it can't resurface likely relevant material while I'm writing about it. @wandcrafting and I wrote about this in part 2 of roamgames, but the conceptual framework could be applied to many data models. https://t.co/wezshrwwaW ([View Tweet](https://twitter.com/RobertHaisfield/status/1492686555659440132)) - I put "AI" into quotes because it doesn't have to be machine learning. It could look at the structure you already have and surface relevant material. If your thought processor has relationships, it can surface surprising information. https://t.co/0uMYsXBpxW ([View Tweet](https://twitter.com/RobertHaisfield/status/1492687541270581252)) - This frame is from @ibdknox - when you resurface past notes/data entry during an exploratory search (brainstorming, drafting, gathering), it should surprise you. Serendipity comes from seeing something new, not sifting through "yeah I know that already" https://t.co/Lpok48fMw7 ([View Tweet](https://twitter.com/RobertHaisfield/status/1492689419035234308)) - There are tons of benefits to showing old material as you're creating new material, as @gordonbrander writes about with search-or-create in Notational Velocity. I will caveat that sometimes wikilink autocompletion distracts me from my thought. https://t.co/zAXQLUqCcg ([View Tweet](https://twitter.com/RobertHaisfield/status/1492690244470120453)) - Distraction is appropriate sometimes. Geists are an interesting one from @gordonbrander. Little spirits that distract you in pleasant ways. Like a conversational assistant. They watch you, ask provocative questions, and resurface possible answers. https://t.co/kxhaQXgE3x ([View Tweet](https://twitter.com/RobertHaisfield/status/1492691257809768448)) - Talking with @devonzuegel recently, she said that most of her synthesis happens in conversations w/ others. They ask surprising questions that force you to reconsider ideas, rotate them, or examine them through a different lens. Surprise is a key ingredient for synthesis. ([View Tweet](https://twitter.com/RobertHaisfield/status/1492693076908732416)) - Imagine: While thinking out loud in your notes, and you hit a block, know what goes in there, but can't quite articulate it - You could describe what goes into the hole, and be presented with suggestions. It could even suggest queries to help you fill it! https://t.co/I5Tl8sy34h ([View Tweet](https://twitter.com/RobertHaisfield/status/1492695308517609472))