It's cool how you can just do this now in 2026. I hope it gets cheaper and easier to do with other big projects written in outdated or just not good enough languages
HN have very primitive comments layout that gives too big of an focus to large responses and first most upvoted post with all its replies. I think just because of that it's better to do something about large responses with little value. I'd rather they just share conversation link
"the current platform direction and climate combined with an algorithm that prioritises hatred, polarisation, and sensationalism, alongside growing privacy and data protection concerns"
While I agree that this platform has a lot of hateful people, it's definitely possible after some basic internet hygiene end up with nice recommendations feed especially after forming solid following list with good people, no hatred no politics only coolest people doing cool things. I like it there. It's the place where things happen that you read about on other cites only weeks later in some twisted form
What examples of AI integrations annoy you? Because I have such wonderful time randomly discovering AI integrations where they actually fit nicely: 1) marimo documentation has ask button to quickly get some help, kind of like way smarter RAG; 2) postman has AI that can write little scripts that visualize responses however you want (for example I turned bunch of user ids into profile links so that I could visit all of them); 3) Grok button on each Twitter post is just amazing to quickly get into what post even references and talks about. 4) Google's AI Mode saved me many clicks, even just Gemini that can quickly fetch when certain TV Show goes live and make reminder is amazing
post trained models strongly inclined to pass response similar to what got them high RL score, it's slightly wrong to keep thinking of LLMs as just next token predictions from dataset's probability distribution like it's some Markov Chain
Job offers require experience in technologies that you won't ever need building solo project. I'm not surprised when those big scale technologies get shoehorned into small project for the sake of learning, showcasing "look I know that one" etc.
Only totally missing this point could explain why someone would make this hyperbole rant page
Yeah but it started being really annoying when you import something like Xray photo. Like chanting "sorry human as LLM I can't answer questions about that" and then after few gaslighting prompts it does it anyway but now I have to take in count that my gaslighting inputs seriously affect answers so way more chance it hallucinates...
Check out their YouTube channel where they show plenty of interesting features. But just to list some I can think of:
- optional reactivity (i.e. you create chain of cells where editing 5th cells in the past causes update down the stream, pretty neat when working with dataframes). Its reactivity is a very cool feature once used to but you might not want it for something like running heavy ML training task so it can be toggled off
- you can switch notebook to multi-column notebook mode
- notebook is a web app that has sidebar with a lot of menus, there cool sections like Docs, Packages (you can download new packages right away there with uv), plenty of LLM integration with their custom prompts where you can reference dataframes so that it would be able to understand schema, some SQL and other DB integrations as well, cells can even contain SQL instead python code and output query result into python variable
- thanks to reactivity it got a lot of interactive elements like sliders buttons text fields or ability to create entire own widgets, there's even mode where all code blocks get hidden and you're left with complete app
- you can make web export of notebook that will translate python to WASM and publish it as fully working static page (though publishing something heavy complex like torch probably won't go well), this fits well with previous point as you can basically build simple interface hide all the code and publish it (like imagine matplotlib with couple of sliders)
- DataFrames (pandas/polars) displayed as interactive tables where you can filter by columns, scroll through pages of rows etc
- notebook stored in a .py format, unlike .ipynb with its json like structure. So code is very Git-friendly but you don't store computation results anymore
GIMP doesn't have "Patch tool", it doesn't have "Inpaint Brush" that is like content aware fill but brush, it doesn't have content aware fill either. I remember there were some plugins that provide features kinda similar to those tools but it was a huge pain to install and use them through nested filter menus. It's a great program for free open source, but it's decades behind on feature parity with other editors, and launch times long enough for me to prefer opening Photopea tab in browser instead
This thread is pure doomerism. Program became free and people made up tens of scenarios why this is bad, awful and literally the end. I mean I understand it's "too good to be true" and some "rug pull" is coming, but people acting like it's "over over".
So far what I've got is a new amazing photo editor that's unlike GIMP has "Inpaint brush" and "Patch tool" and even has free built-in image segmentation model and tons of other good features and even, seems, should support Photoshop plugins. Even if every single new feature going from now on will be gated behind some absurdly priced subscription I'm still totally fine with what i've been given already.
No way they bet entirely on AI thing popping off, unless they'll come up with their own "Nano Banana" idk who really needs simple "text to image" model, and even if they do it's still not that good to be worthy of subscription. On another note subscription right now also gives access to three local running models: upscale, depth map generation and colorizer, pretty cool small models in addition to segmentation that's given for free, I guess they might end up adding additional later there. Another thing I see is that several third party services baked in, for example Dropbox in linked services and two stock image providers Pexels and Pixabay, maybe that covers some cost for them. On mysterious ploy to gather all users data for Big AI I see one concerning thing: Right away program asked whether I want to share data and I selected option "no" but then in settings suspicious "Send usage and performance data to Affinity" was enabled, either I forgot that I pressed yes or it's a bug or something intentional.
this concept has huge potential when done slowly with caution, but I bet this one ends up too politically biased one day, so far seems like the most blatant copy paste of wikipedia texts
Okay if author so ignorant and doesn't even see difference and just genuinely hates this topic why even make an article? "Look see this new things someone else exited about, I don't give a damn about it so much that I'll write post about that, I'll even completely ignore figuring out what's different about it myself and just spew some surface level ignorant overview"