Seems like YT is mostly focusing on the interaction between streamer and audience (which makes sense because streaming is big with young ppl rn).
I understand "ambient streams" to be more like a setting for a group chat or chat room, where you're interacting with friends or strangers only, there's no focus on a single creator/streamer. Like hanging out at an interesting location instead of a featureless room.
> I have a ton of idea about how ambient live streams are the Next Big Thing and how to lean into that.)
Are you me?
It's crazy that these giant screens spend most of their time as black rectangles when they could be windows to the world (with very tasteful/quiet/no-motion advertisements).
Happy to pitch to Netflix, Roku, or other streaming services and/or TV manufacturers. :-)
I think GP meant that the ambient content could be accompanied by a “tasteful” static sponsored message. Imagine a black stripe at the bottom that said “Sponsored by Home Chef”
Given how much ads on the web escalated in intensity though, I don’t think anyone would buy sponsor spots like that though.
We understood what the GP was thinking in terms of ads. The objection is that any form of advertisement, regardless of how tasteful it might be presented, is the polar opposite of the intended purpose of tranquil windows into nature.
People like these kinds of live streams because it offers a relaxing form of escapism. Dumping even a static ad only serves to add distraction from that atmosphere and thus undermines the very point of having the live stream open in the first place.
I think I agree with you tbh. Then again I go pretty far out of my way to not see ads, other than billboards and maybe those on the sides of buses and bus shelters! To me it’s worth paying not to have them. So I bet a screen would be no different for me.
My old Sony TV, just a cheap model, had a (custom for Sony) “Lofi Girl” ambient screensaver it would go into if you configured it to, with quiet music. It was one of my favorite features.
If you're interested in the multiplayer cursors + cursor chat, my philosophy is that every web page deserves to be a place, and pages should feel busy if lots of people are there
I think that’s a very cool philosophy but unfortunately it makes your website unusable for me on mobile. It feels like ants are crawling all over my screen.
First time I saw this, I didn't realize these were other people cursors, I really though the point was to be an annoyance (some people actually do this!). I thought the movement was random, and that it was a way to make it harder for people to see their actual cursor.
The country flags make it much more clear. Still an annoyance, but now I get it, and I am more than amused than annoyed now.
Anyways, it doesn't work for me on Firefox 128.11.0esr, Debian 11 (yes, outdated, I know, but still in LTS), I had to use Chrome/ium for the cursors to show.
It's very cool, but the motion is also incredibly distracting while reading. It literally makes it physically difficult to read.
Might I suggest quickly fading the cursors out entirely as soon as the user starts to scroll, maybe? Then you could have the effect at the start, but be less distracting while reading.
Or just a floating counter in the corner to say how many people are currently viewing it, maybe with the two most popular flags and your own flag and a fourth "other". Because it's one thing to know it's busy (cool, it's popular, I'm participating in something!), but it's another thing to feel busy, distracted, claustrophobic.
I assume you want to prioritize people reading your actual content over the feeling of busyness.
Thanks. Honestly, given all of the random comments in bubbles and other UX I never even saw it. I realize it's obvious once you've seen it, but everything else was so distracting I genuinely didn't. And I was literally trying to just "tune everything out" and read, not assuming there was an option to disable it that I should look for.
Also, "quiet mode" doesn't suggest any clear link to the cursors. I get what you're saying with it, but I'm more likely to assume it's got something to do with the site playing sound effects I don't hear because I've got my device on silent mode.
This is such a cool effect but I'm afraid its novelty is conflicting with its affordances.
Have you seen cases where people are using it in a more familiar manner? Like, they've moved on from the newness but there's still a bunch of people? Feels like being in a subway station where everyone is bumping into each other right now instead of just sharing the space as needed.
yeah this is definitely over the edge -- signal is lost in the noise
my main enjoyment has been to hang out on my own blog (which it generally pretty quiet) and say hi to people as they drop by. I've had a few pleasant interactions that way, and a couple people said hi in Unoffice Hours (link in the left column) or on the socials after
but generally I feel like "ambient togetherness" is just the beginning of something, and it needs to be paired with something more persistent to be useful (like a discord only open to subscribers, that kind of thing), and I haven't gotten around to building that side of it yet.
It helps in my case. Bash is all about text so you just see text being output and manipulated. Something like lisp is about lists and eval, so you mostly see those and equivalent expressions. With something like React, there’s this tree representation of state that is always on your mind.
It’s a more abstract representation than code. Not as fixed, but helps greatly with designing the next steps.
It seems to be a side effect. It doesn't help me in any obvious way. However, when it's happening then I've also ceased to be conscious of writing code. It just flows of its own volition.
I mean, I know I'm writing code, but I'm not consciously seeing, analyzing, formulating, or typing in code.
Edited to add: On reflection, the geometric vision does actually help me sometimes. When there's something wrong with the code I've written (be it a syntax or logical error), then the shapes react in a way that I'll call "dissonant". I pay attention to that signal and more consciously analyze what I've just done, to find the error.
The reason I ask: there's a story about a physicist who was so kinaesthetic in his thinking that somebody walked into his office to find him rolling around on the floor, trying to embody rotations from the point of view of a particular system or something...
I can't say that my own subjective feelings while coding are so useful! But I like to imagine that they're a meaningful contributor to my "taste" of what good/bad looks like
I certainly don't get as deep as that physicist, but I do think I understand. When I'm working out a difficult logical/data/programmatic problem, then I use my hands in a similar way. I'll form hand shapes that in my mind represent geometric shapes and physically use my hands to "arrange" them in various ways until something "clicks" in my head. That "click" isn't the win of finding a solution, but more like a feeling that I have now grasped the true essence of the problem.
Essence of the problem is the right word. If we go with F. Brooks’ words, it’s like all the accidental complexity fade away and all that’s left is the idealistic view of how things should work. Then all that left is making it happens with code
(Outside the UK, the "Listen" link doesn't work except for the most recent ~350 episodes. You'll need to get the program page link instead which is in the footer.)
My assumption is that models are getting cheaper, fast. So you can build now with OpenAI/Anthropic/etc and swap it out for a local or hosted model in a year.
This doesn't work for all use cases but data extraction is pretty safe. Treat it like a database query -- a slow but high availability and relatively cheap call.
While it will become cheaper, it will never be as fast / efficient as 'just' parsing the data the old-fashioned way.
It feels like using AI to do computing things instead of writing code is just like when we moved to relatively inefficient web technology for front-ends, where we needed beefier systems to get the same performance as we used to have, or when cloud computing became a thing and efficiency / speed became a factor of credit card limit instead of code efficiency.
Call me a luddite but I think as software developers we should do better, reduce waste, embrace mechanical sympathy, etc. Using AI to generate some code is fine - it's just the next step in code generators that I've been using throughout all my career IMO. But using AI to do tasks that can also be done 1000x more efficiently, like parsing / processing data, is going in the wrong direction.
I know this particular problem space well. AI is a reasonable solution. WHOIS records are intentionally made to be human readable and not be machine parseable without huge effort because so many people were scraping them. So the same registrar may return records in a huge range of text formats. You can write code to handle them all if you really want to, but if you are not doing it en masse, AI is going to probably be a cheaper solution.
Example: https://github.com/weppos/whois is a very solid library for whois parsing but cannot handle all servers, as they say themselves. That has fifteen + years of work on it.
I think you’re both right, and also both are missing the point.
Using LLMs to parse whois data is okay in the meantime (preferably as a last resort!), but structuring the data properly in the first place (i.e. RDAP) is the better solution in the long run.
Requesting that people think before transferring mission critical code into the hands of LLMs is not being a Luddite lol.
Can you imagine how many ridiculous errors we would have if LLMs structured data into protobufs. Or if they compiled software.
It's more than 1000x more wasteful resources wise too. The llm swiss army knife is the Balenciaga all leather garbage bag option for a vast majority of use cases
Still, I wouldn't use an LLM for what's essentially a database query: by their very nature, LLMs will give you the right answer most of the times, but will sometimes return you wrong information. Better stay on a deterministic DB query in this case.
As usual, arguments for LLMs are based on rosy assumptions about future trajectory. How about we talk about data extraction at that point in the future when models are already cheap enough. And in the meantime just assume the future is uncertain, as it obviously is.
Any policy suggestions need to show how they'll urgently remedy the serious issues with energy, housing, and infrastructure detailed in the report -- or make the case that this issues don't matter.
see my other post with the full-viewport waterhole, that was what I used to get rid of YouTube chrome.