My fiancé works at Anthropic.
asserted
fiancé → work → Anthropic
At the beginning of April, the prominent AI researcher Andrej Karpathy tweeted out an idea that was, for a certain kind of productivity nerd, an infohazard.
asserted
that → tweet → nerd
He described the idea this way: “Something I’m finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest.”
asserted
I → describe → interest
He proceeded to describe his process: adding source documents to a local folder and using an LLM to extract and organize their contents into a Markdown wiki that gets updated as he adds new material.
asserted
he → proceed → material
Karpathy’s AI-related pronouncements are closely followed online — he is, after all, the person who coined the term “vibe coding.”
asserted
who → relate → term
And so, seemingly within hours, the web had filled up with GitHub repos, YouTube videos and Substack posts about how to set up an LLM wiki for yourself.
asserted
web → fill → yourself
I call the idea an infohazard because, simply by becoming aware of it, I had ensured that I would devote the next several weeks to building it, without having any idea whether it would benefit me at all.
asserted
it → call → me
As it so happens, it has.
asserted
it → happen → ?
Of everything I tried this year to get better at the deskbound parts of my job, the LLM wiki has easily been the most useful.
asserted
wiki → try → job
Like a vintage sports car, it requires a fair degree of maintenance, and there are almost certainly easier ways to create a personal knowledge base.
asserted
it → require → base
But if you do any sort of work that has made you crave the help of a good research assistant, an LLM wiki might be worth your time.
uncertain
wiki → do → assistant
The wiki is the centerpiece for my annual productivity post, where I run down any changes to the way I work that might be interesting to others afflicted by an inexplicable enthusiasm for software.
uncertain
that → run → software
So with that, here’s what I’m still doing from last year, what I stopped doing, and what’s new.
asserted
what → ’ → what
What I'm still doing
Given how often I switch apps, the best test for whether something actually makes me more productive is longevity.
asserted
me → do → apps
Do I install the app on a new machine?
asserted
I → install → machine
Do I renew the subscription when it’s time?
asserted
it → renew → subscription
Can I point to the places where it actually saves me time?
asserted
it → point → time
Three apps I’ve recommended in previous years still clear that bar.
Raycast, a launcher app that replaces Spotlight, remains my preferred way to navigate a computer.
asserted
that → recommend → computer
When I press ⌘-space, the Raycast window instantly materializes and lets me perform actions across most of the apps that I use.
asserted
I → press → that
I look up words; I do math; I reposition windows; I access my clipboard history; I open websites.
asserted
I → look → websites
And thanks to a paid upgrade, I do tons of simple AI searches in the Raycast window.
asserted
I → pay → window
(I use GPT-5.5 Instant here for the high quality-to-speed ratio.)
asserted
I → use → ratio
When I’m done, there’s no getting lost in a jumble of open tabs or windows — Raycast simply fades back into the background.
asserted
Raycast → do → background
At this point, I really can’t imagine my Mac without it.
asserted
I → imagine → it
And in a nice development since last year, it’s now available for Windows as well.
asserted
it → ’ → Windows
Capacities, which bills itself as “a studio for your mind,” is where I keep my daily journal.
asserted
I → bill → journal
Each morning, I write a bit about whatever’s on my mind.
asserted
whatever → write → mind
Then, I add notable news links from Techmeme to the bottom of my journal and tag them.
asserted
I → add → them
The result is that when I’m writing a story or preparing for a podcast, I can click a tag like “Labor” and instantly see all the stories I’ve saved on that subject since I started building this system a couple years ago.
asserted
I → write → system
This has been enormously helpful in planning our current podcast miniseries on AI and productivity, since we discuss jobs news on each episode.
asserted
we → plan → episode
I used to spend a lot of time digging through databases or running fruitless Google searches in an effort to jog my memory about something I had read; Capacities ensures that it’s all just a click away.
asserted
it → use → something
You could easily do this with any number of apps, but after three years I still find myself appreciating the simplicity and calm of Capacities.
uncertain
myself → do → Capacities
Finally, last year I mentioned testing an app called Recall that (anticipating the LLM wiki!) helps you save and organize content from the web.
asserted
you → mention → web
I found myself using it less for that purpose over the past year, except for one remaining killer use case: its Chrome extension provides near-instant text summaries of YouTube videos.
asserted
extension → find → videos
Over the past year, I’ve saved myself many hours by dumping podcasts I feel like I should listen to for work into Recall and simply skimming the summaries.
asserted
I → save → summaries
What I stopped doing
The two things I stopped doing over the past year are related to each other — and to the LLM wiki.
asserted
I → stop → wiki
For years now, I’ve been seeking a solution to problems of memory.
asserted
I → seek → memory
I’ve been a tech reporter for almost 16 years, have written a newsletter for almost nine, and have written this newsletter for six.
asserted
I → write → six
For much of that time, I’ve been publishing stories and saving research materials in various places.
asserted
I → publish → places
And when a news story comes along that draws on some of that history, I want to find that context as quickly as possible.
asserted
I → come → context
…and 92 more, not listed.
The Aporia
