💾 Time to close this one up. Saying goodbye to my Ghost S1 build named pequeno. I think I’ll name this one chonky. Next up, fan curves.

💾 Audio books are not my thing. I do listen to podcasts though, so I figured I’d give non-fiction audio books a try again. The challenge is “highlighting” for future reference. Just found that Snipd supports uploading own files (DRM-free) and it’s amazing! Snipd → Readwise → Obsidian 👌

💾 Ouch, now was not the time for my RAM to start bugging out. 💸

💾 People dream of multi-user support on iPad… I don’t think macOS has its shit together either. I constantly run into annoying issues with the shared home laptop. E.g. external drives mounting for the wrong user, networking issues (VPN or other weirdness). No problems with a single user. #endofrant

💾 Excited about Obsidian introducing Bases core plugin! I’m hopeful it will replace Dataview in my vaults since my use cases are simple. As a rule, I aim to keep the list of community plugins as short as possible and rely on built-in functionalities.

TLDR; Wasted a day on trying out local LLMs as reflection assistants for my journal with underwhelming results.

As the year turned, I’m still reflecting on 2024 and contemplating how to approach 2025. I’ve kept somewhat regular journals during the year. Reading through the weekly reviews has been very helpful in processing the last 12 months. Neverthless, there is a lot of information and I didn’t go through each and every daily journal. In addition, I’ve logged highlights from other sources such as blog posts, Youtube videos, etc.

Since I was anyway inspired to review my setup (I use Obsidian on mobile and desktop), I decided to explore local LLMs as potential reflection assistants. Privacy is important to me, therefore the emphasis on locality. Having had reviewed my journal manually already, I knew what a high quality bar of analysis would be.

Unfortunately, I ended up sinking a good chunk of my day off (Ascension is a public holiday in Finland) for not much benefit. At least I did learn something about local LLMs and a few apps. In fact one of them, LM Studio is user-friendly enough and I’d keep it on my machine for more tinkering.

One of the first hurdles is the ability to ingest a number of files, i.e. my Obsidian vault. The first option I tried was NVidia’s Chat With RTX. In principle, it supports exactly that. Except that it doesn’t have official support for .md files. Luckily, there was supposed to be a simple tweak to cheat it into ingesting them. However, it seemed it would just ingest 3 files and I couldn’t figure out why. As a side note, I had lost my patience debugging its installation via the NVidia app failing, when the the installer from the website worked out of the box.

Next I moved onto the already mentioned LM Studio. It was really straight-forward to set up and get going. It does support uploading up to 5 files, incl. some size limitation. So I resorted to Co-Pilot writing a script of merging my Obsidian vaults files into one .txt file, this went swimmingly. I used the default Llama 3.2 1B model. With some prodding I did get some meaningful answers, but it was clear as day I can’t trust it. Still I’m up from trying out other models as it’s so easy to obtain them.

The final attempt for the day was with PrivateGPT.dev. It’s also supposed to be able to ingest a number of files. As it turns out, you can’t specify a parent directory. This one required more of a setup and I was trying to be careful with the Python environments on my machine. After some fiddling downloading dependencies from left and right, ultimately the web UI ran… and failed miserably importing my files. I threw the towel.

In retrospect, I’m disappointed with this experience, but there is still some curiosity left in me to fiddle more with this. In principle, analyzing my writing for insights should be an excellent application for LLMs. Alas, my first take was a resounding failure. At least I did learn a few things and got to do something technical on my day off. Now… let’s try passing this post through LM studio for proof-reading.

P.S.

Assistant llama-3.2-1b-instruct

Spelling and Grammar Mistakes:

• TLDR" should be spelled as “TODAY’S SUMMARY”

I can’t.

💾 Year 2024: Outlook Calendar still doesn’t support “Cancel future occurrences”. The workaround is to update the invitation with an end date and confuse everyone. One should not change history. Old man yells at cloud.

💾 Many exciting and deeply technical news from WWDC… But you know what simple usability improvement I’m excited about - native window tiling in Sequoia, with keyboard shortcuts. We’ve suffered for so long. Rectangle did wonders but it had its hiccups.

💾 There is more to artificial general intelligence, and my own human intelligence, than putting the most likely word in front of the other. Yes, it’s a reductive comment.

We needed an SD card urgently and just bought a cheap one from a basic department store. I should know better, as I’ve worked professionally with SD card benchmarking and filesystems… but still thought “We just need a few photos, how bad can it be?”. Turns out - really bad.

The card got corrupted after first use, refused reformatting and repartitioning (only 2/32GB were useable). I managed to salvage the photos from this first photo session - total luck. Btw, not all repair tools are made equal. The macOS built-in fsck didn’t do the job, but the one from dosfstools did the trick.

Anyway - don’t buy the cheapest card you find from your department store, if you care about what you keep on it that is.

💾 Readwise Reader has a new feature of note. I’ve been looking forward to it so much, that I’m almost a bit anxious to check it out. Posting this and heading over to the web app.

📁 RSS Folders on Web – You can now organize your RSS feeds into folders and pin individual feeds to the left sidebar resulting in a more traditional feed reading experience.

💾 There is some friction for me entering text in my Obsidian notes across devices. I’m dialling back to basics to see if I’ve overcomplicated it myself.

• No line numbers
• No vim mode
• No Live Preview

In the process, I’ll also lean in the macOS built-in shortcuts for editing and navigating text.

💾 While there is disappointment about platforms striking training data deals with AI vendors, I think we see the writing on the wall. If companies don’t make such agreements, AI will be trained on public data anyway as it’s been so far. At least formal deals work against the fair use defense.

💾 Conversed a bit with ChatGPT about GPU rendering pipelines. Now that I’m testing Pro, figured I’d ask it to generate a diagram as well. DALL-E has a different “brain”. Welp, I guess it makes for a nice T-shirt design… from afar.

💾 I’m using Proton VPN and I’ve noticed a lot more often sites just don’t work. And it’s not Proton’s fault. Simply everyone everywhere wants to track the cr*p out of visitors. Just today couldn’t access the city library and a movie festival sites. Come oon! 😡

💾 Once upon a time, I was interested in writing reliable and efficient shell scripts. Using shell builtins was an eye-opener. Today I remembered about parameter expansion:

mystr="$(uname -o)"

# If we find "Linux", we get an empty string
[ ! -n "${mystr##*Linux*}" ] && printf "We are on Linux.\n"

💾 Kudos to @mikehaynes for the excellent Outpost theme! Just the refresh I was looking for, and it was so easy to customize based on my previous theme.

💾 How I miss Things. It’s truly a usability masterpiece. Decided to drop it for 2 reasons - not cleared at work (especially the cloud bit); too separate from my notes and writing. Imagine if they entered the editor space and integrated the tasks alongside notes seamlessly. One can dream.

💾 Streaks is an amazing app. Every NY I return to it (for obvious reasons 😅). Its super power is simplicity. Super impressed with the team who manage to add more continuously, while keeping complexity in check. I’ve now slotted Streaks snuggly in my daily workflow.

💾 Micro.bloggers (Cc: @help), what is your workflow for testing Micro.blog theme changes with a test blog and a Github repo? I thought of playing with the test blog on a branch and using main for “production”. But I don’t think I can point the test blog to a repo branch via the Micro.blog UI 🤔

💾 Groundbreaking - customers not buying kills companies.

“What this advertising boycott is going to do is it’s going to kill the company, and the whole world will know that those advertisers killed the company and we will document it in great detail,” Musk said.

Source: Ars Technica

💾 Sometimes I’d like to engage with online communities on specific topics. Discord feels too overwhelming and noisy (inherent to chat tools, I guess). Reddit, in theory, should be a sweet spot - but again so many distractions. Matrix, IRC, email - feel too clunky. 🤷🏻‍♂️ Digital introvert problems.

💾 Captain’s Log, Stardate 2023.8. I used a FAT-formatted USB stick as a last resort to make the office printer cooperate.