CrypticShift

I'm using dynalist.io. I'm just used to it. The app is Electron, though. but I feel it is responsive enough for writing.

I have an observation on the original use of OPML, i.e., for outliners (the one I utilize frequently) : Many recent outliners (logseq, roam) added custom properties/fields (for DB-like block-level queries), but they don't seem to support them when importing or exporting opml. The only "field" that is somewhat supported is the "_note" [1].

[1] https://web.archive.org/web/20131205072742/https://jeffreyki...

I agree. The original reason [1] for the gemini name seems artificial for a generic chatbot. It is OK for the model, and I'm sure a lot of "work" was put into "validating" it for the assistant, or... was it?

[1] https://the-decoder.com/how-googles-gemini-ai-model-got-its-...

While today is about Gemini Advanced and its new capabilities, next week we'll share more details on what's coming for developers and Cloud customers [1]

[1] https://blog.google/technology/ai/google-gemini-update-sunda...

I'm confused: Is this GPT-4 or GPT-3.5? The last reasonable answer I found (I forgot where) mentioned that they only downgrade it to v3.5 during peak hours. Still, even if it is v4, it seems to behave differently. The copilot should somehow be added independently to the LLM leaderboards.

> Very often there is an implicit discretization

Not just on the temporal dimension. I find that what is called "pitch bend" is highly relevant to the "quality" of most music (even western popular music).

You can use a custom range on hn.algolia.com [1].

[1] https://hn.algolia.com/?dateEnd=1707004800&dateRange=custom&...

The article does not discuss two-wheelers, although I assume they are included (?). Can one imagine a ban imposed solely on internal combustion motorcycles ?

Could this be a simpler first step for populous third-world countries? I mean, they are quite popular there (= real impact), while also being much cheaper (= easier to replace).

Someone should make a censorship/alignment (whatever you want to call it) benchmark for LLMs.

There are some userscript filters you can customize : https://greasyfork.org/en/scripts?q=hacker+news+filter

Not a filter, but this one "bypass" paywalled links https://github.com/MostlyEmre/hn-anti-paywall

On a related note, with the (tens of) thousands of "custom GPTs" coming up in the next few years, it would be interesting if the chat would automatically recommend using any one of them in response to a particular query. In a way, it is as if it is directing you to a (human-made) better engineered (pre) prompt.

Phind [1] uses the larger 34B Model. Still, I'm also curious what they are gonna do with this one.

[1] https://news.ycombinator.com/item?id=38088538

Keyword Filter List -> https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Keyword Filter Alert -> https://alerthn.com/

AI Tagged List -> https://histre.com/hn/?tags=+lisp

AI Tagged RSS -> https://www.kadoa.com/hacksnack

Make sure to use GPT-4 (not 3.5).

I would say 10k/1k+ is too much for pocket. I prefer bookmark managers that have a dual folder/tag system. I try to take advantage of this duality in my taxonomy choices: hierarchical folders are at the base, with tags as more "lateral" descriptors.

Good Point.

Even animals "self-trancend" for their family/herd. it is instinctual.

Most Humans "self-trancend" for their tribe/nation. it is emotional (+ instinctual).

But it is not really "self-transcendence" as posited here: universal and mindful (to not say "intellectual").

Did you try ChatGPT? It may also assist you in the matching.

As a visual helper, you could also chart the 1K tags as embeddings in a 2D graph [1]

My setup when manually working on this is a split-screen outliner (filter, drag&drop...).

Once the taxonomy is done, I assume you will not be able to reflect those levels back in pocket ? I just see a flat list of tags there.

[1] https://cookbook.openai.com/examples/visualizing_embeddings_...

There are many new tools emerging. Here is a raw list. Some are still alpha. Most are not free. And I believe only some of them specifically parse/import social media links.

https://mymind.com/

https://betterstacks.com/

https://fabric.so/

https://allclues.ai/

https://sublime.app/

It is true that Bard/ChatGPT is just two clicks away. But never underestimate the power of defaults. This is definitely not a good default for writing anywhere on the web. Google could at least have made this an extension instead.

I've yet to see a serious comparison of grammarly alternatives.

I feel there is already a lot of potential for using ChatGPT to replace most written language tools. For some actions, like rephrasing, I find that some prompt engeneering is necessary to precisely modify my "text" and/or get a useful "commentary" on it. As you said, iteration is also key. Just make sure to use GPT-4 (not 3.5).

You can do it in hn.algolia.com -> https://hn.algolia.com/?dateRange=all&query=by%3Aohjeez&sort...

To use this app's (peculiar) descriptor, it is more "pleasant" for those who deal with a lot of text "information". This includes (but is not limited to) students, academics.... At the core level, it is about using many "doc files" without juggling between multiple windows, file managers, etc. Then, in addition, some may feel the (real?) need for other features. You can look at this extensive list [1] and see if you feel you are "missing" any of them, although feelings could be deceptive :)

[1] https://noteapps.info/features

In a way, this reminds me of the (much more ambitious) system of resolvers [1] of the (now defunct) tomahawk player [0]

The idea was you just give it the metadata and it "resolves" it into any service. I really like this idea. it kind of lives on in "playlist converters" like tunemymusic or soundiiz. but it is not the same as it being built into the player itself (like spotube albeit with a different more straightforward aim here)

[0] https://github.com/tomahawk-player/tomahawk

[1] https://github.com/tomahawk-player/tomahawk-resolvers

This is "Chinese". I personally don't mind, especially if it is open source. But I've seen another China (and electron) based note-taking app vilified for this reason.

FYI It is Siyuan [1]. What is unique about it is the logseq-like block-level granularity (ref/embed/query) and (soon) the notion-like databases.

[1] https://github.com/siyuan-note/siyuan

It seems this is only for the new S24 series. I hope this gradually extends to lower ranges. Personally, I find a 2023 phone at 1/8th of the price [1] to be sufficient for my use (spec-wise)

Now a removable battery and Less "freezing" [2] please.

[1] https://www.gsmarena.com/compare.php3?idPhone1=12583&idPhone...

[2] https://news.ycombinator.com/item?id=38995719

So the author is keeping notes in a flat hierarchy and then using the graph.

He stated that the goal is "greater understanding," and then he just (in all honesty) states that the nice graphical topology represents my learning journey and keeps me motivated.

That is exactly the problem: motivation and a history of learning do not always lead to "greater understanding." On the contrary, I find that our 'notes' are often transient while our internal understanding morphs subtly yet swiftly (especially when learning something new).

That is why I think 'organic linking' works better for very large AND more definitive content. Wikipedia is the best example.

For personal use, I prefer more structure. The outliner/database combo (like logseq) is the best tool for that.

And Frankly as long as I can easily download my data in a readable format, I'm OK with any tool. It does not have to be built as plain text files.

But I do like Obsidian :)

> We don't even understand the universe.

Exactly... Only when we (better ?) understand the universe will we know if (and how) we (meaningfully) add much to it.

So this is the new generation of Javascript self-hostable wikis inspired by Notion. I remember in the late 2000s, most were PHP-based, with Twiki being an exception: it was Perl-based and also had a powerful wiki/database combination [1]. But it was not intuitive to use. That's what Notion nailed a decade later.

Now All the new (OSS ?) Notion clones like Outline are still deficient in that second database aspect. I hope we are eventually getting there.

[1] https://twiki.org/cgi-bin/view/Codev/StructuredWiki

You can spend unlimited time working on yourself. I see no unhealthy unbalance in that as long as you have no inner need or outer pressure to produce (or you are able to control both).

Personally I find it tricky, though. Am I doing nothing for this long because I am just getting lazy (-> need resolve) or depressed (-> need real help)? It does not seem to be your case, so just take your time. Take your time.