> I can tell you're not a parent, because if you were, you would know that basically none of the digital solutions provided by tech companies to facilitate gating adult material from children actually work or are in any way thoughtfully designed.
I can tell you Nintendo's parental controls work correctly. Even the eShop doesn't display any content not suitable for younger ages for accounts under the parental control restrictions if configured correctly.
> I have never seen any form of create generative model output (be that image, text, audio, or video) which I would rather see than the original prompt.
I've used LLM before to document command-line tools and APIs I've made; they aren't the final product since I also tweaked the writing and fixed misunderstandings from the LLM. I don't think the author would appreciate the original prompts, where I essentially just dump a lot of code and give instructions in bullet point form on what to output.
These generated documentation are immensely useful, and I use them all the time for myself. I prefer the documentation to reading the code because finding what I need at a glance is not trivial nor is remembering all the conditions, prerequisites, etc.
That being said, the article seems to focus on a use case where LLM is ill-suited. It's not suited for writing papers to pretend you wrote a paper.
> I say this because I believe that your original thoughts are far more interesting
Looking at the example posted, I'm not convinced that most people's original thoughts on gimbal lock will be more interesting than a succinct summary by an LLM.
> I thought that was the whole point of ActivityPub?
To allow others to see your posts, yes.
But if your instance goes away, instances can't fetch those posts anymore and admins of Mastodon instances shutting down tend to run a process that prunes users/history across the fedi too. Even if they don't, that information fades, especially as status authentication is a thing now (checks original instance for the post existence).
Then further, if you move to another server, that history is not attached to your new account either etc.
> And why is your instance shutting down?
This time it's due to new laws that the UK has introduced around social media websites. Previously it's been stuff like admin didn't want to do it after getting bullied. Not enough money. Lost all data and can't be bothered anymore etc.
> I guess "had" is the underrated word there.
I don't believe he's on Twitter anymore.
> Sure, but I' wondering if that need will be met anywhere on the net.
I believe this is currently handled by "algorithms" on the major social media networks.
> The fediverse stuff works pretty well as far as I can see.
Not the grand parent, my instance is shutting down in January. I have no way to migrate my posts. This is the sixth time I've gone through this over the years. I don't feel like dealing with this stuff anymore.
> The only thing the fediverse lacks that Twitter really provided was the starfucking celebrity culture.
If you pay attention to the few notable people in their fields, they describe some of the problems such as:
It's interesting to me that the ones who really have a problem with this change are those who already left Twitter. But, having removed themselves from being the target demographic and already indicated that they won't return; I'm very much in agreement with you here that these changes are now much more with the flow.
This article is so bizarre compared to my personal experience.
> If you’re on Twitter today, you’re likely experiencing an app that is more frequently broken
I'm not, I used to get loading issues all the time in years past, not anymore.
> more random
More random ads, but actually less stuff I don't want.
> and more unhinged
The day after Elon sent all those people home to let them know if they were fired, I never got the political stuff anymore, just the content I was interacting with all along, which happens to be photography and art.
> Many users are complaining about seeing more content in the new For You feed that they don’t want to see.
I scrolled through mine, literally only photography stuff if you ignore the ads.
>
The problem is, Musk’s plan doesn’t seem to be working as well as he’s saying. Twitter’s web traffic dropped by nearly 8 percent last month compared to the year before, and has been dropping for the past three months year over year, according to new estimates from data intelligence firm SimilarWeb. This directly contradicts Musk’s claims that usage is up.
I don't think that's a contradiction because similarweb doesn't measure app usage.
My photography interactions have only increased since last year too...
> The Pixel, though, in my opinion, produces pretty bad photos.
I've had the opportunity to play with iPhone 14, it has 12 megapixel cameras, but I find that the image quality is no difference to 8 megapixels of detail, and if you use third party applications that can take pictures without post processing, it's often closer to 6 megapixels and the grain in the image is not great. Smart phones don't appear to be resolving the detail advertised.
> I shoot in RAW
Shooting raw on my Android phones (since the iPhone 14 won't let me) shows there is clearly very little dynamic range, which when trying to do beautiful processing on skin tones or such, come out very flat (regardless if going for natural or unrealistically perfect).
I do a lot of photography in poor lighting, I over expose a lot so I can the details in the shadow, and then bring it back down in post processing in raw. Try to do anything similar on phones, terrible grain. Leave it to the phone processing, it does some really bad approaches at AEB with post processing and not very good handling on the HDR merge, plus the exposure time is huge, meh.
> But I suppose that my eye is good enough to know real bokeh from fake.
It's not convincing to me because you can see stuff like the background between strands of hair in focus. I imagine it's only going to get better at that.
The biggest problem with phone photography is sensor size currently and it seems unlikely we're going to have larger sensor sizes when it requires more flange distance for the optics.
Current Apple chargers and iPhones support USB-PD. The phones only fast charge properly with MFI. This is literally looks more like they swapped the connector and did no other changes. Also, from my reading of the common charger directive, this doesn't breach the rules.
I'm not the grand parent, but your comment stuck out to me.
> I don't think that's fair as developers spend ages customising everything anyway
I've got into development. I use KDE Neon now and really, the only thing I customize on a new install is left single and double click defaults in the OS, changing system performance mode to prefer performance over battery and install the IDE and Yakuake for a better terminal.
Other than that, the starting development for me is literally just installing the IDE, cloning the repository and opening the project file (and the IDE will pull in all the dependencies, linter settings automatically).
> I can open a terminal instantly and start writing instantly.
One of the best things about Yakuake I love is that it's literally that. Hit F12, multi-tab terminal slides down like the old quake console (yes, that's where the name came from).
> The idea is to have a view built by the connections people on your instance make. The view from tech.lgbt is different from the view from photog.social, and that's a feature.
They look completely identical, the feature isn't working.
I'm sceptical it's even an intended feature as much as just the overall result of the protocol design not putting too much thought into it.
> Smaller nodes can subscribe to relays if they find it too limiting.
I looked into it, that's actually a better solution than the old one I was aware of which involved a bot account automatically subscribing to different accounts across the fediverse. It's my understanding that needs to be explicitly configured by a Mastodon administrator for a subscription though, out of the hands of the user.
This still isn't particularly user friendly and I still think this is a terrible user experience, particularly when comments like these appear:
>>>> Hashtags play a huge role on Mastodon - discover people by following hashtags
>>> I'm probably twiddling the wrong buttons or something but whenever I've clicked on a Fediverse hashtag, it's always just been the hashtag on the originating server. Is there a way to get a wider view that I've missed?
Need to have people on your home node following people on the other node for your own node to be aware. Really unfriendly to small nodes / people who run their own personal nodes.
Which isn't that obvious that there is a problem as a new user. People are posting all kinds of trash on Twitter equally, you don't expect to find the entire node is blocked because some users are posting things some others disagree with.
> Depending on your interpretation of your local laws, you might pretty much be obligated to defederate from that sort of thing.
Not that I'm defending this, but because you mentioned it; I looked it up. Mastodon.social isn't obliged because it's located in Germany and so is Eugen; so this reasoning doesn't apply. Block the problem users, not the entire instance.
> I think focusing on what has the biggest userbase or can give you the most reach is kind of missing the point with regard to the Fediverse.
I looked for which node was ran with reliability and scaling built in, because I didn't want to be on a node that is ran by some hobbyist that is just going to give up one day, I also don't want to be on a server that falls over because scaling problems. Pawoo was top of the list.
This was further validated when my second account on another node disappeared one day because the owner decided they didn't want to continue running their node anymore. And further validated on other nodes where the server became so slow it was practically unusable, the server owner didn't really have any idea how to deal with the userbase.
I tried running Mastodon myself, but quickly found it awful to keep updated; at the time (I don't know if it still does), it wasn't very clean to update and was fairly manual. I eventually switched to GNU Social with the Qwitter addon because it was easier to maintain and update, only for Mastodon to decide to drop support for the federating protocol it was using one day.
You're absolutely right though, I don't get the key point of the fediverse currently.
> Without investors or stock prices or the need to sell ads, a lot of the motivation behind that sort of thinking evaporates.
If we take your logic to its ultimate conclusion, you're saying that without investors or stock prices or the need to sell ads, the motivation behind making a good and sane experience evaporates, considering the state it's been in for years. I honestly wouldn't try to disparage open source community projects like that to be honest.
> If you're looking at Mastodon and thinking "this is worthless, I can't build my brand with this," good, you're right, it's built that way on purpose.
I don't care about stock, brands. I'm looking at Mastodon and thinking that it's got serious communication, usability, data and privacy issues. It has so many basic things fundamentally broken. On Twitter, if you delete a post, it's gone. On Mastodon, delete a post, well, it's deleted on your node. But the rest of the fediverse has replicated it and it's not deleted. The other nodes aren't even keeping your data maliciously, it's just poorly designed and the fact it still won't tell you that "deleting" messages doesn't really "delete" them is plain vindinctive at this point after it has been raised as an issue repeatedly.
> The concept of copyright was created in response to the development of the printing press. It was a reaction to a disruptive technology.
Absolutely, one of the major factors was that it allowed individuals to benefit directly off someone elses work without having made substantial changes. The protection was intended for the original works it self and derivatives too close to the original content.
> We are now at the threshold of a new disruptive technology that is likely to bring about profound economic changes in the arts.
This already happened with photography taking over portraits and tracing, the response wasn't to outright ban it, or really prevent it either. When technology made photography more accessible, to the point it was going to be disruptive to professionals in the field, the response again wasn't to outright ban it, or really prevent it either. This is despite the fact that it has litterally destroyed a significant amount of jobs to achieve conviniences that we now all enjoy.
I feel like the AI issue is a parallel to above situation. People are now given better tools to generate/create art themselves and as long as it isn't blatant copies, derivatives too close to the original content, it probably should be have similar rules in my opinion.
> It only functions by drawing from their works.
You can train AI models by taking photos and then vectorizing/toonifying/paintify etc. depending on what you're aiming for with various wildly available non-AI filters. Stylistic ideas are possible to implement into these filters, I have some experience having done so with making plugins for processing my photos. So, that isn't even a strict requirement for generation. So, even in the case where you ban AI from learning from people made art (even in the situation where they would allow it), there are ways to still train the AI models regardless to achieve a similar result.
There is another problem that hasn't been discussed, enforcement is going to be a very interesting problem considering how international borders for information/data are virtually non-existent now and it's becoming relatively difficult to even distinguish if a piece was generated by an AI or by a person. The economic changes are likely coming in regardless from my point of view. It's going to be either people are using it illegally if banned regardless or people using it legally if it isn't -- I just do not see this changing either way.
From my personal experience in using Mastodon, the node operators basically have different ideas about what is acceptable or not and black/white list different nodes based on these ideas rather than leaving it up to the user to decide who they want to see posts from.
Unfortunately, a lot of this is really unclear when you're first trying to dip your toes into the fediverse and it gets really confusing in some threads when you're missing some of the messages.
I think the worst part is that they've previously advertised you have control of your data. But I can tell you that if you accidentally post something on Twitter and want to delete it, it's trivial, it's gone. On Mastodon, tough, it's been federated and now your deleted message lives on across multiple nodes and you can't delete it -- not out of maliciousness, but because of oversight in the federation protocol design.
The reason I bring the above up is, again, you can miss a lot of bits of conversations entirely dependent on which node you are because some messages are now deleted on the origin node, but not the nodes that has other people replying on.
I can tell you Nintendo's parental controls work correctly. Even the eShop doesn't display any content not suitable for younger ages for accounts under the parental control restrictions if configured correctly.