Mobile offline sync is a lost cause. The dev environment, even on Android, is so hostile you’ll never get a good experience.

Joplin comes close, but it’s still extremely unreliable and I’ve had many dropped notes. It also takes hours to sync a large corpus.

I wrote my own web app using Axum and flask that I use. Check out dokuwiki as well.

Oh good to know.

It used to be awful but I’m glad to hear it’s improving.

Maybe snapdrop?

When I was obsd I did FTP and rsync for everything. Syncthing had dinner performance issues for me.

Maybe Seafile but I had a bad time with that.

From memory MTP is pretty flaky and quite slow.

ADB push is pretty good but at that stage rsync is just as easy.

Put SSH in the phone and you can do it all from the computer too.

Are you able to buy unlocked directly from Google? I typically avoid the carrier when I can.

Yeah I did the smart thing and walked away from research.

Total dead end, Industry don’t care at all.

To be fair, wireguard is pretty painless.

Absolutely that’s what the internet was made for!

But family photos keep a bit more secure, Particularly if it’s syncing directly from your phone, I take a lot of explicit photos of my wife, but also code that I’m writing on my computer, or the kids playing, etc.

I don’t think I would have made too much of a difference because the state-of-the-art models still aren’t a database.

Maybe more recent models could store more information in a smaller number of parameters, but it’s probably going to come down to the size of the model.

The Only exception there is if there is indeed some pattern in modern history that the model is able to learn, but I really doubt that.

What this article really calls to light is that people tend to use these models for things that they’re not good at because it’s being marketed contrary to what it is.

I think they all would have performed significantly better with a degree of context.

Trying to use a large language model like a database is simply A misapplication of the technology.

The real question is if you gave a human an entire library of history. Would they be able to identify relevant paragraphs based on a paragraph that only contains semantic information? The answer is probably not. This is the way that we need to be using these things.

Unfortunately companies like openai really want this to be the next Google because there’s so much money to be hired by selling this is a product to businesses who don’t care to roll more efficient solutions.

Well, that’s simply not true. The llm is simply trained on patterns. Human history doesn’t really have clear rules such like programming languages, so it’s not going to be able to internalise that very well. But the English language does have patterns so If you used a Semantic or hybrid Search over a corpus of content and then used an LLM to synthesise well structured summaries and responses, it would probably be fairly usable.

The big challenge that we’re facing with media today is that many authors do not have any understanding of statistics, programming or data science/ ML.

Lllm is not ai, It’s simply an application of an NN over a large data set that works really well. So well, in fact that the runtime penalty is outweighed by its utility.

I would have killed for these a decade ago and they’re an absolute game changer With a lot of potential to do a lot of good. Unfortunately the uninitiated among us have elected to treat them like a silver bullet because they think it’s the next dot com bubble

Yeah man, why not, update here I’d love to check it out!

Maybe I should upload some content too :)

Content, no clue how to fix it though.

• Starting Strength book
  • The program is simplistic, dont follow all the advice in the book but the guidance around compound lifts is good
  • Follow that basic program for a month or two (Or ice cream fitness), depending on your level of fitness you might like PHAT or PHUL.
• Couch to 5k
  • Once you Can run a 10K It’s simply a matter of determining what your goals are

On the software side:

• Fitnotes
  • Although it’s proprietary, it’s free and that allows you to export your data into SQLite. Any llm (even a local one) will cobble together a dashboard with plotly/jupyter or even PyQt without much pain
• Waistline Is an application available on F-Droid, This is open source. So, send the developer a donation if you like it.

Well yeah, self-instruct is how a lot of these models are trained. Bootstrap training data off a larger model and fine-tune a pre-existing model of that.

It’s similar but different.

On some devices with Linux suspend can still consume a lot of power, I’ve had some pain with this in the past with Void but runit boots quick so non-issue.

I suppose another perspective is encryption, when the device Is powered off. It’s going to be encrypted so there might be an extra degree of security there.

When I was performing dart analytics and teaching at the same time I would turn off my machine between classes just in case. But I still wanted it to boot fast because I’d have to then go and teach.

https://en.wikipedia.org/wiki/Lightship

Oh, that is a fair point. We did have a donor kebab place, Which definitely blow Burger King and McDonald’s out of the water.

But variety is nice and in the absence of that there was nothing other than KFC and Burger King.

A lot of areas don’t have other restaurants. It’s a sea of fast food that can only be driven to.

City is different, but before I had the savings to move, I u didn’t have much choice

It’s also just been a tough period to grow up. Depends on region but housing, did and cost of living in many have been significantly harder than in the past.

Every year is harder to get by than the last, despite unprecedented advances in science and tech.