Articles

I only just found out what these embedding LLMs do the other day, and Google just dropped a new 308M param model!

I really like how they’re open-weighting(?) these smaller models that can run on super low end hardware. They dropped Gemma 3 270M a few weeks ago and whilst “dumb”, they’re very easy to fine-tune on places like Google Colab.

Pretty excited to see how apps could use these on-device models for search features:

enables you to build applications using techniques such as Retrieval Augmented Generation (RAG) and semantic search that run directly on your hardware

I’ve been thinking about this on and off since OpenAI released the improved memory in ChatGPT. I subscribed to ChatGPT plus for a few months during my exams, so it could help me with the more niche topics or sort my notes, etc., but the memory was actually very impressive. You could start a new chat and it’d have a very good recall of what you discussed previously. This is probably not that surprising. I have no clue how it actually works on a technical level; however, it’s probably not too difficult?

But this is a huge lock-in for LLM providers. It’s a very competitive market, and there are different factors pushing and pulling users between different platforms. Parker Ortolani also blogs about this and for me this issue is highlighted here:

I have been attempting to use them in my daily life but ChatGPT’s extensive knowledge of me, my interests, my current life situation, and the projects that I have been working on with it have created a great deal of, you guessed it, friction.

I was surprised to see huge companies like OpenAI, Claude and Cursor (tons of others too!) using the open MCP spec to allow their models to access data from other places. This results in a much improved UX for pro users.

This is sort of unrelated, but I think the openness of this protocol is quite impressive. It works well (we can clearly tell it does as it’s still heavily used in this fast-paced space) and yet allows people to use the latest models of their choice, whilst accessing the same information.

I think memory could be like this. Unsure how it would work in a secure and encrypted manner, but it’d be really cool if there was some similar equivalent for memory. You could plug your saved memory into any given LLM and it’s got all your past context right there on the fly.

The only issue is it’d probably be much more advanced for the average person sending ChatGPT a request - although I guess you could say the same for MCP.

This may come across as really dumb as I’m not sure how to do it, and I’m not sure if this is already being worked on or talked about, but I think it’ll be important. It’ll have to be easy to setup too, whilst also somehow working reliably. A secure, self-hostable way to bring in prior knowledge.

<img src=“https://jackk.micro.blog/uploads/2025/chatgpt-memory-e1744306989473.png" width=“600” height=“374” alt=“A pop-up window titled “Introducing new, improved memory” discusses ChatGPT’s ability to remember past chats, with options to learn more, dismiss, or view details.">

I’m curious how much charge I could save on my iPhone if I didn’t rely on it for everything. I might charge my iPod later to use that for music. That along with using my real camera over the iPhone one… hmm 🤔

Realistically, I won’t use the iPod instead of my iPhone bc you need to have songs on it before, but I’m already using my Canon PowerShot camera instead of my iPhone camera in most cases, so we’ll see!

Update: You can find the code for this on GitHub (feel free to fork!): github.com/j4ckxyz/T…

I’ve been reading my news on Tapestry more for the past ~2 days since I’m off Bluesky for a bit (primarily used that to read haha) so I vibe coded (ye, not me) this little Tapestry Connector to use Nitter to show Tweets!

Currently just provides a nicer way to add Nitter RSS feeds, but not exactly sure how to get profile pictures of the accounts to show in the feed itself 🤔 but open for suggestions!

Not sure if there’s meant to be some “official” way to share these connectors, but I’ll be sharing the code and latest release on GitHub. You can go to the “Releases” tab and install it into Tapestry. From there you can just add a new connector, type in the username, and it should just appear in your feed!

From testing on an active account (ok Elon being addicted to Twitter can be useful apparently) it doesn’t seem to show ALL Tweets, nor does it seem to show as SOON as it’s been Tweeted, but it’s a good start for now, at least until someone smarter might make a better one.

so I know Pixelfed is likely getting a new wave of users atm across their instances, but I think now is the PERFECT time for someone to make an Instagram app based on atproto.

Mastodon didn’t really work after Elon purchased Twitter, and Pixelfed is the same layout. But Bluesky has a large amount of users (~10M MAU) - so it has a chance to catch Instagram’s user base imo!

Once people figure out their new single online identity (atproto) it also offers new developers a whole pool of new users they can already bring in to their applications.

Hope someone does this! Tons of discussion about it right now!

#atproto #bluesky

Little post about Mastodon/Fediverse’s general view on #Bluesky.

I think it’s a bit far to say it’s all centralised, etc.. and a lot of people who are saying stuff like this don’t even know how atproto works. Sure, a few do, and there are pros and cons to ActivityPub and atproto, but looking through my friend Kuba’s Bluesky feed which shows Mastodon/ActivityPub posts mentioning Bluesky, you can really see a ton of negative opinions.

Of course it’s your account and instances, so post what you want (within the rules, duh!!) but it’s not right to say it’s billionaire owned, because last time I checked (2 seconds ago on a quick Google search!) there’s no single billionaire working for Bluesky. Sure, they have investors, but they’re interested in the protocol itself, not Bluesky the app. They have clearly stated this - many Bluesky engineers have.

Just putting my say out there. If you really want people to join Mastodon I wouldn’t be so narrow-minded and almost not letting people use other apps. It’s their choice – CHOICE, your fav word 😏

Rant over 💬