daily sentences 2408122225
daily sentences 2408122225
on the status of all the apps making information work possible
OneNote, I logged in 2007, Evernote in 2012, and Notion 2016. All I reset in 2019 to align with {annual_observation}s. Notetaking tools are material ecosystems which synthesises information before it becomes knowledge. After 2022 the tool set required realignment to accommodate generative AI. 2023 is about text and images with attempts to disrupt search. 2024 is video, sounds, agents and artifacts. Claude and ChatGPT do not compete but start becoming complementary tools. Opensource models, Llama and Mistral are better. No one is convinced that AI will increase productivity as much as it is hyped. 2025 to project, is the year of devices, AI integrated into operating systems.
To work with AI, there is first implementation and then its application. To implement is to decide if the application in external or internal. Defining use case is paramount. What is the problem the services solve is urgent to articulate. Auditing the status of existing workflows is the step to follow. Workflow as an ideology is a post generative AI era phenomenon. Picking the right tools for every component of an assembly line takes trial and error. Time is lost in testing these combinations out, while keeping in check how efficient the automation process is. Locating optimal personalisation is the pursuit of the times.
It is possible to draw an extended writing workflow, but numerous the parts, the more tools creep in. Evernote and Notion are sites of material logging and planning. Plans leads to outlines which are built on. Content edit is where OneNote is introduced. Unlike in the past when everything was file based, here only the final edited version of the text is created. Based on the dissemination requirements, text is processed further. Proofing is file based. LLM application for text based workflows is one aspect. Planning for accompanying visuals is an addition to elaborate on. Log, outline, edit and proof, as states of work is in principle transferable to image making.
Creative workflows has never been optimal. It is always messy and resource intensive since investment is not adequately quantified. Efficiency of creative work is a state to attain. {x items with y resources, against z time} is easier said than done. It is a habit, and therefore a discipline to form. Monitoring consumption is the task. What are the distinct sources ideas are drawn from, what are their types, and in what quantity. Evernote, Notion and OneNote presents different states of information process before finalised as a document. From scraps to an outline and then stacks, every processing state makes for more refined iteration of a document under production.
References & Links
Devs Need System Design Tools, Not Diagramming Tools. Navigating The XY Problem: Reframing Questions to Solve the Right Problems.
Change blindness. 21 months later [Post]
I don’t think anyone is completely certain about where AI is going, but we do know that things have changed very quickly, … If this rate of change continues, the world will look very different in another 21 months. The only way to know is to live through it.
How I Use "AI” [Workflow]
Replacing my Right Hand with AI [Workflow]
AI is a Hot Mess [Paper]
How should you adopt LLMs? [Essay]
How the design of AI tools turns writing into a “collage”. Exploring UI fragmentation in the design of new text tools. [Workflow]
10 Developer Productivity Boosts from Generative AI [List]
40 AI prompts to boost your marketing team’s creativity and efficiency. Learn to use AI Prompts as a marketer [List]
So, you want to be a programmer? [Video]
How Sprinting Slows You Down: A Better Way to Build Software. Sprints promise to accelerate development, but often do the opposite. Check out this alternative approach to building software. [Workflow]
Decision Making Frameworks [Essay]
The New Internet [Essay]
