And you don't have to pay for it? The lower tiers that Kagi offers doesn't have AI stuff it seems.

The lower tiers don't have Kagi Assistant, but they do have the Quick Answer and Summarization features. You can turn them off or not use them though, so it shouldn't really be an issue.

I find it interesting that another comment states that they love Kagi exactly because Kagi uses all their search terms as opposed to Google. Not doubting your experience here, but I wonder how those experiences can differ so much.

Say something positive about kagi, receive karma.

Say something negative about google, receive karma.

No place with this kind of system should be relied on for veracious claims.

It's not a system, you are just describing sentiment, people generally like Kagi and generally dislike Google.

> people generally like Kagi and generally dislike Google

The reason the words "on hacker news" is missing from that statement is the karma system. I am describing echo chambers.

No way that people on a tech-focused forum like a technology designed for power users! Must be an echo chamber!

Hey! Thanks for asking what I would do to make it better. First, of course, the karma system has to go - scoring can still exist without being visible and therefore pressuring people to be performative. In addition to that, I would adjust the rankings to show not only the most popular posts but also some of the most controversial ones (controversy being defined as least agreement among voters). These are changes that will alienate many, however, and unlikely to be incorporated by a platform that amount to a content marketing tool for an investment group.

That comment you refer to confused me, as I have a similar experience as OP (posted in a sibling comment).

If you hire someone to so a job for you, you are accountable. Unless they bring their own car and accept liability.

Honest question: what is not modern about JupyterLab? I know JupyterLab has existed for a long time, but continuous development has kept it modern.

My take:

- The UI is over bloated and bugged, sometimes things scroll, sometimes they don't, sometimes you have to refresh the page. You cannot easily change the UI as lots of CSS parts have hard coded fixed sizes.

- The settings are all over the place, from py files in ~/.jupyter to ini files to auto generated command line parameters.

- The overall architecture is monolithic and hard to break down, jupyter proxy is a good example of the hacks you have to go to to reuse parts of jupyter

- The front end technology (Lumino) is ad hoc and cannot be reused, I had to write my own react components basically reimplementing the whole protocol, come on its 2025.

- The whole automation around nbconvert is error prone and fragile

This is mixing quite a few different things (backend, frontend, auxiliary CLI utilities).

No time to write a lengthy reply here, but I think it's worth separating legitimate like-for-like comparison with a wider feeling on the ecosystem.

The need to start the server is really annoying. Especially when you have notebooks in multiple places, or multiple virtual envs.

This is why I moved to working with Jupyter notebooks in VS Code, there is no server to manually start.

Vscode will start the server for you, in practice. This is great if you just want to get going. It gives you a bit less flexibility though, if you want to do something fancy.

Vscode can also connect to existing servers. This can be very useful. For instance, you can put a ton of data and CPU in a server and work with vscode on a small laptop. If network latency is low enough, this works great.

I can't read this either.

Edit: at a distance it's easier to read

If you squint it’s easier too. I wonder if lowering the resolution of the image would make the text visible to ocr.

I wonder if you could do a composite image, like bracketed images, and so give the model multiple goes, for which it could amalgamate results. So, you could do an exposure bracket, do a focus/blur, maybe a stretch/compression, or an adjustment for font-height as a proportion of the image.

Feed all of the alternatives to the model, tell it they each have the same textual content?

I'm not sure if you realize that this is _not_ a per user price, but a flat monthly fee. To me, that seems insanely cheap.

If you have a small team of 3 people who get paid 4k gross (severely under paid in all likeliness) then Jelly is 0.3% of your total cost.

As a European, I don't get this. You frame it as a choice between two people: a badass vs a bumbling idiot. Why should it be about the personalities of the candidates and not be a choice between the kind of policies that are being proposed and the ability to reason about them?

While I concede that subjectively, Trump can come across as badass and Harris might have been "phony" at times, when it comes to the policies both sides are defending during debates and even the way they express themselves verbally during debates, Harris comes across to me as someone who has better ideas and can explain them far better.

I think that the difference between politics in Europe vs the US is that the latter is based much more on cult of personality. And sadly that mode of operation is becoming more common here in Europe as well.

I personally don't care for the personal attacks and mud slinging. Let's bring back intelligent debates.

I made some generalizations, but if you want to discuss policies, these policies I believe are vital to the future of our country and the main reason I voted for him.

I approve of Trump's policies:

* A secure and stable border where we do not allow millions of illegals to cross into our country

* Abortion states issue to deal with

* Smaller government - hiring of Elon musk to eliminate thousands of Federal jobs

* Drill baby drill

* End homelessness for American veterans

* The free American college education through an online school

* Bringing medical production back to America

* The establishment of tariffs on Chinese made products to reflect China's tariffs on American imported goods

* Ending the wars

Really neat solution to build a nice looking resume! You get bonus points for using JSON Resume. I think it's smart to encode your resume in a machine readable format. That way you can generate any presentation you want from the pure data.

That also answers the question so many people seem to have: why not write HTML/CSS by hand instead of using Eleventy. That doesn't work if you want to separate data from presentation

The website states that Reactive Resume is open source, but no matter where I look on the website, I cannot find a link to the code.

Can you provide that?

Also question: does Reactive Resume let you import and export in JSON Resume format?

It does allow JSON import/export format, and it used to even allow more granular customization of css before it got a rewrite. Here’s the source code: https://github.com/AmruthPillai/Reactive-Resume

Caldendly/cal.com don't compete with/replace CalDAV. They solve different problems. CalDAV is a protocol for exchanging calendar data. Calendly and alternatives are tools to let people outside your organization schedule an appointment with you.