That's such a broad criteria that it's next to impossible to give good recommendations. What does that mean? Could you make a list of requirements that would make one platform more "creator-friendly" than another?
I think one of the biggest key questions is, how much do you need/want the platform to assist with authoring, engagement, and discovery?
YouTube is okay at this; TikTok is excellent at it. TikTok provides the fyp, community management, authoring and captioning tools, access to thousands of sound/music clips to use where they're handling the legal side, etc. Ostensibly this is very, very "creator-friendly", but only for a fairly specific niche of TikTok-like content, and only if you're not skirting the line of what they'll censor.
Vimeo has always felt to me more like a larger Vidyard— the content is 100% on you, but they'll handle the technical side as far as encoding and distributing, and give you a clean player widget you can embed on your own site.
Embedding a HTML5 video player with some javascript on top is quite trivial. I think the actual, reliable, fast and somewhat affordable global distribution of video streams is the actual meat of the problem from the perspective of someone who would eventually just self-host.
That doesn't mean that a considerable amount of non-tech or low-tech (for certain definitions of low) people wouldn't rely on the actual tech expertise in terms of encodings, poster frames, subtitles, chapter marks, time-coded comments or similar things.
Sure, but that comes back to who the "creator" is in this scenario and what their actual needs are. Embeddable video hosting might be perfect for an e-learning site, or an indie film studio, or someone providing paid online music lessons, or whatever.
But if the "creator" in this scenario is khaby.lame or kallmekris, a service that helps them host their own videos on their own site is of zero use; those are social media personalities who take the hosting and technical side for granted— their primary need is exposure to a gigantic audience.
For which Vimeo was never the golden choice before. Most people I know who pay for vimeo are film makers who need hosting, sometimws for piblic stuff, sometimes with a password for preview purposes.
I maybe didn't phrase it well, but I'm talking about companies that make just enough money to sustain themselves but never quite grow nor go out of business. (I founded one and it wasn't very much fun. We eventually killed it, but it was never unprofitable). Anyhow it seems like floatplane hasn't made much of an impact - it's not huge like Youtube or growing massively - so does it have long term prospects?
That's because they don't charge content creators; as it's a pay to play only platform, they make money from the individuals that subscribe to you. The amount you charge is up to you, though they only list the minimum once you have an account.
Let's rephrase that, if you have easy access to hosting and [sufficient] bandwidth, can anyone recommend creator-friendly tooling such that you can just host your content yourself?
I'm so done with relying on third parties to do a poor job hosting stuff I created :(
Hmm, if you go the the instances list on joinpeertube.org, and hit "display" or "no opinion" on the "Sensitive Videos" filter, you don't have to go down more than a couple before seeing awful white nationalist content instances :|
I don't think this is that relevant. It's like saying the web is full of shit. It's true, but should we throw the baby out with the bath water?
I host a PeerTube instance for my choir and I don't quite care about the existence of PeerTube instances hosting questionable content. I didn't enable federation though. Still hesitant to do this.
Perhaps I should have added more to my comment. I love Peertube and it is very valuable software. I just question if the joinpeertube.org website shouldn't blacklist some of those from appearing on their website. Having such links there definitely wont increase the likelihood of your average content creator using the platform.
Ah, so they should blacklist things you don't like. Well, while they're at it, there's plenty of stuff out there I don't like so let's blacklist that as well. I'm fairly sure there are things out there which neither you nor I would want to blacklist but which should be banned according to others who don't care about our blacklists.
Maybe the current system works after all? Just don't visit those instances you dislike, this is what I do. Give people the option to choose for themselves whether they want to partake of material which you or I or my neighbour or his dog finds offending.
I was just suggesting that they not link to white nationalist content their website. The website that is the first result of searching for "peertube" in duckduckgo or Google.
What is the problem with what they keep saying? They don't want to send their would-be viewers to a site that promotes white nationalist content on their home page. Who would, besides racists and edgelords?
Exactly how does dialogue with white nationalists help the rest of us grow? Their goals are to remove, silence, or destroy people that have different immutable characteristics. How do you suppose that impacts free expression?
Don't you realise the same can be said about many of those categories I mentioned elsewhere in this thread? For sure those communists only are out to topple out liberal democracy, why would we support that? Those AFA-anarchists are only out to do the same using even more violent methods, not shunning anything to create their dystopia, who would want to support that? Why should people who call out for the eradication of all men be allowed to use Peertube? Who wants to watch black supremacist material other than black supremacists and why should they be allowed to use this tool for their nefarious purposes? This list can be made as long as you want, eventually asking why the knitting club should be allowed to spread their racism using Peertube [1].
So, no. Either everyone is censored or no-one is. I vote for the latter, let the law decide what is allowed and keep it at that. If it isn't forbidden, it is allowed.
They don't have to. Their site will be entirely separate and hosted under whatever domain they want.
Imagine using WordPress to construct your website. If some page existed that listed all the sites that use WordPress, and that listed some disagreeable ones (as it surely would), that has very little impact on your usage of the software.
I don't care what's on PeerTube's front page any more than I care about what's on YouTube's front page (ie. not at all), since I always just search for what I want.
Viewers don't install anything. That would be the content creator here because the question is "as a content creator who wants to self host, what can I use?", but that could also be someone the content creator knows, or someone providing a PeerTube instance for content creators.
PeerTube is a piece of open source software meant to build a YouTube-like platform, with optional federation (and P2P to allow servers to offload some bandwidth to viewers, but I don't know how well this works in practice with the widespread asymmetrical or mobile internet connections). It is not for the viewer to install.
Internet video is stupidly simple these days, now that the Flash requirement is long dead: export your video in H.264 and WebM to cover all of your bases (Firefox, Chrome, Edge, Safari, and iOS Safari all support H.264, WebM is supported by Firefox, Chrome, Edge, Safari on MacOS 11.3 or later, Safari on iOS 15 or later and Android Chrome.) Then you can just upload it to an S3/CloudFront bucket and slap a <video> tag on literally any web site. It's basically as simple as serving images these days. https://developer.mozilla.org/en-US/docs/Web/HTML/Element/vi...
Most of your garden variety blog/CMS engines have plugins that can even automatically take media uploads and stash them on Amazon or a similar CDN. For something that's not even that big, that might even be overkill. If the amount of data transferred monthly is low enough, you could probably just serve it off a VPS.
This is probably not quite enough in many cases.
You may need a search engine that indexes your videos, a comment section, a way to refer to sections of the videos, something that provides several versions / qualities so the video can be viewed on any internet connection but still be available in high quality when this is possible, adaptive streaming, and many other things I'm not thinking of. All that in a user interface optimized for videos.
You can probably implement a part of these features imperfectly in a regular CMS and with the video tag, but you may as well use readily-available software that do this job for you well.
If you host, occasionally, one or two videos, that's probably the best thing to do, but if you are to create a channel, you are probably better of using dedicated software.
And of course, people always forget about accessibility for those with disabilities. Everyone making their own half baked video solutions creating a more user-hostile web for the disabled.
I mean you can just throw more cloud services at the problem* and use AWS video transcription. It works fine. About as well as MS and Google automated transcriptions.
* I think it’s dubious that S3 & CloudFront is really DIY anyway but hey, if you already broke the seal.
That's a very primitive approach. Once you're dealing with adaptive formats (DASH, HLS) containing multiple renditions (different bitrates) and/or codecs, it gets more complicated. I get that a single h264 codec/mp4 (webm) container file is easy enough as it is widely supported but once you're dealing with larger video's and getting them served over different connection types/speeds, it gets trickier. Atleast with services like Vimeo/YouTube, you don't have to bother too much.
Well, that's what they sell. Vimeo charges reasonably for transcoding and storing all these different versions of videos. YouTube does it for free to stifle competition and stay the only place advertisers can go if they want a captive audience (with TikTok and Facebook, both also free, being their only real competitors).
If I had to actually recommend a service it'd be Cloudflare Stream[0], which allows you to upload videos, CF transcodes them, and you get a link to a page with just the video player[1], and they offer options like only allowing embeds on certain origins and disabling/enabling MP4 downloads. The pricing is fairly reasonable as well at $1 per 1,000 watch minutes and $5 per 1,000 minutes of stored video.
>Then you can just upload it to an S3/CloudFront bucket and slap a <video> tag on literally any web site.
You know how much cloudfront bandwidth costs? $0.085/GB. Vimeo has a 2TB/month limit before they "may reach out to discuss a more suitable plan for your needs". If you consume that much bandwidth, it would already be costing you $170/month on AWS for the bandwidth alone. I suspect viemo's plans are still going to be cheaper.
One disadvantage is that this forces less technical users to learn about technical details of media, which itself can get very complex.
For example you seem to know your way with the solutions you propose, but I already see an issue, as in reality WebM is a format limited to VP8 and VP9, and cannot contain H.264 video.
But Chrome ignored this limitation and wrote a muxer that is actually able to store H.264 inside WebM files, thus going out of spec... which means different tools might have different degrees of support for those files.
Arguably, a former user of YouTube or Vimeo shouldn't have to learn about all this, and shouldn't have to end up reading bug reports such as that one (and this is just scratching the can of worms that is all the topic of codecs, encodings, containers, etc.)
If you have easy access to hosting and enough bandwidth, a <video> tag and static HTML is "good enough". If you want to get really fancy and are willing to touch JS, you can have some code swap out sources based on the screen size, if the video buffer has run dry recently, etc.
The killer problem is specifically the bandwidth. There is not enough capacity to scale video over IP transit[0]; individual creators will get hosed by bandwidth costs. The way that every video platform gets around this is peering[1], but that requires lots of capital expenditure; and last-mile ISPs will play shenanigans with peering negotiations[2]. This is also why video is a monopoly market: large companies can negotiate way better bandwidth deals than you can.
Someone else mentioned PeerTube; I personally do not consider it a solution because P2P is a privacy nightmare. It's still fairly common still for vexatious litigants[3] to sue members of BitTorrent swarms to try and extort settlements out of people. Not to mention, how the heck do you handle GDPR in an environment where every one of your customers gets every other customer's IP address?
[0] What most people think of when they talk about "buying an Internet connection" - you pay someone else to route your packets.
[1] How ISPs connect to one another. You set up a link between your network and theirs, either in the form of physically wiring up extra cabling or connecting up different switches in a colocation center. That link then carries all the traffic between you and them (and either side's transit customers).
This saves money over transit, if you have lots of traffic between you and that ISP.
[2] Notably, Comcast refused to upgrade their link with Level-3 because Netflix was one of their customers
[3] Also regular copyright holders too, but usually they give up once they get a DMCA takedown processed.
Right now our pricing is $1/user, but with a minimum of $500. We're considering adding a self service option at a lower price point (as this definitely prices out some potential customers), but right now we only offer very white-glove service for our customers and its not economically viable for us to cater to smaller customers with the level of dedication we want to deliver to each of them. We have a dedicated account manager for all our customers, offer iOS and Android apps released under the customer organizations, develop features for the platform based on customer requests, help with migration and design, etc. It's interesting that Vimeo is also now not catering to smaller creators, as its something we decided not to focus on also for viability reasons early on and instead double down on providing a better platform for a higher price point. Vimeo's decision also makes me reconsider if we should allow a cheaper self service option on Instill - the thing I worry about is that given variable costs like streaming we'd have to cover, we'd still need to charge some amount that people didn't feel was "cheap" and so they would expect some greater value to accompany the offering that we wouldn't be able to deliver (ex. we wouldn't be able to have as fast support or an AM help them set up their offering, etc.)
I work at AE, and we've actually been meaning to fix that for a while, but for some reason we prioritized making another theme of our blog that looks just like pg's blog (https://ae.studio/blog/victims-of-vimeo#pg) instead of making links that people realize are links.
I think we were trying to not be too promotional or something, but we should probably fix this, and we will now!
How many of the problems plaguing the modern web could be solved by universal computer literacy -- teaching folks, along with typing and Word/Powerpoint/Excel -- how to host a static website that holds text/photos/video.
Definitely would help a lot. Although I'm not sure we can expect everyone to develop the simplest next step of literacy that creators might need here- having that site also be able to have user accounts, charge users, restrict access appropriately. Since these things are pretty universal, I do also think the no-code space is attempting to meet people halfway there -- letting people do stuff on their own with basic literacy without having to reinvent the wheel each time requiring some deeper (even if still basic) knowledge. Of course some centralized solutions do have predatory pricing to take advantage of the fact that many people need that abstraction layer, but I suspect the problem is also two fold - running the quick math on some of the example creators about to be kicked off of Vimeo, I suspect many of them would find the costs of running this service (especially if they wanted "good" streaming) themselves to be expensive also, not counting the additional time that would be required of them even if they could do it.
I'm not the post creator, but I'm one of the founders of Instill. I don't think it was intended as such as the HN crowd isn't really our target customer (non-technical health & fitness content creators who need a comprehensive / white glove solution to launching a subscription content platform, not people who can engage in a debate about the cheapest way to set up your own infra to stream videos haha). Ev had some interesting thoughts about Vimeo's huge shift which were informed by us having started this platform some months ago based in part on some pushpack from content creators on Vimeo. Of course, as an early startup, we will also take any attention we can get so I don't hate if this wound up being a little bit of both ;)
Digitalocean spaces is $5 for 250 GB while Contabo offers the same for $2.49 / mo. I'm sure there are others too. I think DO also offers a free CDN with edge caching.
If not, just put it behind a cloudflare or amazon cloudfront proxy and you can stream unlimited videos very cheap (probably in 4k i guess). Also for most cases it works best to use the KISS <video controls> tag instead of any silly skins that make users think.
Even hosting your own content, it cannot sustain any reasonable virality. Any widespread consumption requires a giant buffer of Big Tech with deep pockets.
Yes, Reddit is a customer, but they use Data (our quality of service analytics tool), not Video (our suite of video streaming APIs). So...it's our fault if they don't have visibility into you having a bad time, but we aren't in a place to actually fix it quite yet :)
I'm very glad to hear that this is the case, as the developer experience for streaming via Mux looks awesome - much less of a headache than building out a transcoding/delivery pipeline with separate transcoding APIs, object storage, and CDNs.
I'll be giving the Mux streaming API a try for my next project.
Mux is awesome. We are using Mux for Instill. Super reliable and by far the most developer friendly option we found. There is ways to do it cheaper for sure, but when evaluating all the time spent setting up and maintaining all the infra required to do that (especially since we want to get as close to zero downtime, working on almost all device types, fast global delivery, wanted to get Instill up and running fast etc.) we decided it was well worth the cost.
Also funnily enough, Vimeo uses Mux (their analytics offering Mux Data, but still pretty crazy).
The idea here is that there are certainly costs of storage and streaming. And there are certainly SOME customers that probably aren't optimal to serve.
The question is whether one can help those customers grow, and if not, avoid the bait/switch/shock paradigm of Vimeo.
How many domain-specific creators would benefit from a better model that could scale with them?
> How many domain-specific creators would benefit from a better model that could scale with them?
Most of them, and I kind of suspect that most creators already do.
I've been involved in a half dozen different video distribution projects, mostly in the education and non-profit spaces and mostly in the last couple of years (for obvious reasons).
In all but one case, we just threw some <video> tags into a free static site template and hosted the videos themselves directly on the same VPS hosting the static site.
In the other case, a single server probably wouldn't have scaled, so we had... gasp... TWO VPSs. (Didn't even bother with a software load balancer... just did it by hand by giving out different urls to different groups).
Uploading videos to youtube et al. didn't even occur to us. But then, we were hosting content for an audience that was already seeking us out because of who we were independent of the video content, so we didn't need the discoverability aspect.
I suspect this still describes the vast majority of high-quality content hosting, even if the influencer economy or whatever gets the majority of attention. And I'd be surprised if many folks bother uploading those videos to youtube et al
It's like the rest of the economy. There are some behemoths that everyone pays attention to, but a huge fraction of economic activity is a small handful of folks with a truck/storefont/field.
I get the AWS is terrible, charges too much etc. But after being burnt, repeatedly, by unlimited offerings elsewhere, at some point it's just worth paying their storage price or whatever and just know they will be very unlikely to bait and switch you on pricing.
Yes, I'm looking at your unlimited backup storage data plans deals, your "lifetime" offers, your unmetered / unlimited (but wildly oversubscribed and/or we'll give you a call if you actually spin up 20 instances and max transfer on each).
It seems a little bad-faith of the article to relegate to a footnote at the bottom their admission that the practice they're complaining about was changed three days before their article was written.
87 comments
[ 2.7 ms ] story [ 141 ms ] threadYouTube is okay at this; TikTok is excellent at it. TikTok provides the fyp, community management, authoring and captioning tools, access to thousands of sound/music clips to use where they're handling the legal side, etc. Ostensibly this is very, very "creator-friendly", but only for a fairly specific niche of TikTok-like content, and only if you're not skirting the line of what they'll censor.
Vimeo has always felt to me more like a larger Vidyard— the content is 100% on you, but they'll handle the technical side as far as encoding and distributing, and give you a clean player widget you can embed on your own site.
That doesn't mean that a considerable amount of non-tech or low-tech (for certain definitions of low) people wouldn't rely on the actual tech expertise in terms of encodings, poster frames, subtitles, chapter marks, time-coded comments or similar things.
But if the "creator" in this scenario is khaby.lame or kallmekris, a service that helps them host their own videos on their own site is of zero use; those are social media personalities who take the hosting and technical side for granted— their primary need is exposure to a gigantic audience.
A bit like soundcloud for video.
Erm, aren't those exactly opposed?
"Drift along" as long as you are profitable is the definition of sustainable, no?
"Attracting too much audience" is anathema to sustainable if you can't maintain reliability or profitability in the face of the flood.
That's because they don't charge content creators; as it's a pay to play only platform, they make money from the individuals that subscribe to you. The amount you charge is up to you, though they only list the minimum once you have an account.
Let's rephrase that, if you have easy access to hosting and [sufficient] bandwidth, can anyone recommend creator-friendly tooling such that you can just host your content yourself?
I'm so done with relying on third parties to do a poor job hosting stuff I created :(
https://joinpeertube.org/
https://github.com/Chocobozzz/PeerTube/blob/develop/support/...
I host a PeerTube instance for my choir and I don't quite care about the existence of PeerTube instances hosting questionable content. I didn't enable federation though. Still hesitant to do this.
Maybe the current system works after all? Just don't visit those instances you dislike, this is what I do. Give people the option to choose for themselves whether they want to partake of material which you or I or my neighbour or his dog finds offending.
So, no. Either everyone is censored or no-one is. I vote for the latter, let the law decide what is allowed and keep it at that. If it isn't forbidden, it is allowed.
[1] this is not a joke, see https://www.vox.com/the-goods/2019/2/25/18234950/knitting-ra... where they seriously talk about 'Fiber artists of color' (?) pointing out the 'racism' (?) in the knitting 'community'.
Imagine using WordPress to construct your website. If some page existed that listed all the sites that use WordPress, and that listed some disagreeable ones (as it surely would), that has very little impact on your usage of the software.
[0] https://news.ycombinator.com/newsguidelines.html
PeerTube is a piece of open source software meant to build a YouTube-like platform, with optional federation (and P2P to allow servers to offload some bandwidth to viewers, but I don't know how well this works in practice with the widespread asymmetrical or mobile internet connections). It is not for the viewer to install.
Most of your garden variety blog/CMS engines have plugins that can even automatically take media uploads and stash them on Amazon or a similar CDN. For something that's not even that big, that might even be overkill. If the amount of data transferred monthly is low enough, you could probably just serve it off a VPS.
You can probably implement a part of these features imperfectly in a regular CMS and with the video tag, but you may as well use readily-available software that do this job for you well.
If you host, occasionally, one or two videos, that's probably the best thing to do, but if you are to create a channel, you are probably better of using dedicated software.
* I think it’s dubious that S3 & CloudFront is really DIY anyway but hey, if you already broke the seal.
If I had to actually recommend a service it'd be Cloudflare Stream[0], which allows you to upload videos, CF transcodes them, and you get a link to a page with just the video player[1], and they offer options like only allowing embeds on certain origins and disabling/enabling MP4 downloads. The pricing is fairly reasonable as well at $1 per 1,000 watch minutes and $5 per 1,000 minutes of stored video.
0: https://www.cloudflare.com/products/cloudflare-stream/
1: https://watch.videodelivery.net/b9d528d5eece459b80113823cefd...
but why does it need to have all these different formats and bitrates?
varying devices/platforms support varying formats
>bitrates
so some guy on watching on a 3G connection doesn't need to watch your 4K stream.
You know how much cloudfront bandwidth costs? $0.085/GB. Vimeo has a 2TB/month limit before they "may reach out to discuss a more suitable plan for your needs". If you consume that much bandwidth, it would already be costing you $170/month on AWS for the bandwidth alone. I suspect viemo's plans are still going to be cheaper.
For example you seem to know your way with the solutions you propose, but I already see an issue, as in reality WebM is a format limited to VP8 and VP9, and cannot contain H.264 video.
But Chrome ignored this limitation and wrote a muxer that is actually able to store H.264 inside WebM files, thus going out of spec... which means different tools might have different degrees of support for those files.
https://bugs.chromium.org/p/chromium/issues/detail?id=980822
Arguably, a former user of YouTube or Vimeo shouldn't have to learn about all this, and shouldn't have to end up reading bug reports such as that one (and this is just scratching the can of worms that is all the topic of codecs, encodings, containers, etc.)
The killer problem is specifically the bandwidth. There is not enough capacity to scale video over IP transit[0]; individual creators will get hosed by bandwidth costs. The way that every video platform gets around this is peering[1], but that requires lots of capital expenditure; and last-mile ISPs will play shenanigans with peering negotiations[2]. This is also why video is a monopoly market: large companies can negotiate way better bandwidth deals than you can.
Someone else mentioned PeerTube; I personally do not consider it a solution because P2P is a privacy nightmare. It's still fairly common still for vexatious litigants[3] to sue members of BitTorrent swarms to try and extort settlements out of people. Not to mention, how the heck do you handle GDPR in an environment where every one of your customers gets every other customer's IP address?
[0] What most people think of when they talk about "buying an Internet connection" - you pay someone else to route your packets.
[1] How ISPs connect to one another. You set up a link between your network and theirs, either in the form of physically wiring up extra cabling or connecting up different switches in a colocation center. That link then carries all the traffic between you and them (and either side's transit customers).
This saves money over transit, if you have lots of traffic between you and that ISP.
[2] Notably, Comcast refused to upgrade their link with Level-3 because Netflix was one of their customers
[3] Also regular copyright holders too, but usually they give up once they get a DMCA takedown processed.
A typical content creator is unlikely to have enough demand for their videos to congest IP transit links.
IP transit is also very,very cheap. At scale IP transit is only about 5 cents per Mbps.
Source: built and operated a global CDN in-house
Looks like someone built a solution awfully quickly!
Did I miss something or does the blog not link to the new platform?
Took me a minute to find it again because their visited link color is the same as the text color
I work at AE, and we've actually been meaning to fix that for a while, but for some reason we prioritized making another theme of our blog that looks just like pg's blog (https://ae.studio/blog/victims-of-vimeo#pg) instead of making links that people realize are links.
I think we were trying to not be too promotional or something, but we should probably fix this, and we will now!
If not, just put it behind a cloudflare or amazon cloudfront proxy and you can stream unlimited videos very cheap (probably in 4k i guess). Also for most cases it works best to use the KISS <video controls> tag instead of any silly skins that make users think.
(1) https://contabo.com/en/object-storage/
If you are hosting only your own content it's pretty easy. Once you put that upload link on the site it gets a lot more complicated.
https://aws.amazon.com/lightsail/pricing/
https://docs.hetzner.com/robot/general/traffic/
for selfhosting your own videos that should be sufficient.
Yes, Reddit is a customer, but they use Data (our quality of service analytics tool), not Video (our suite of video streaming APIs). So...it's our fault if they don't have visibility into you having a bad time, but we aren't in a place to actually fix it quite yet :)
I'll be giving the Mux streaming API a try for my next project.
Also funnily enough, Vimeo uses Mux (their analytics offering Mux Data, but still pretty crazy).
The question is whether one can help those customers grow, and if not, avoid the bait/switch/shock paradigm of Vimeo.
How many domain-specific creators would benefit from a better model that could scale with them?
Most of them, and I kind of suspect that most creators already do.
I've been involved in a half dozen different video distribution projects, mostly in the education and non-profit spaces and mostly in the last couple of years (for obvious reasons).
In all but one case, we just threw some <video> tags into a free static site template and hosted the videos themselves directly on the same VPS hosting the static site.
In the other case, a single server probably wouldn't have scaled, so we had... gasp... TWO VPSs. (Didn't even bother with a software load balancer... just did it by hand by giving out different urls to different groups).
Uploading videos to youtube et al. didn't even occur to us. But then, we were hosting content for an audience that was already seeking us out because of who we were independent of the video content, so we didn't need the discoverability aspect.
I suspect this still describes the vast majority of high-quality content hosting, even if the influencer economy or whatever gets the majority of attention. And I'd be surprised if many folks bother uploading those videos to youtube et al
It's like the rest of the economy. There are some behemoths that everyone pays attention to, but a huge fraction of economic activity is a small handful of folks with a truck/storefont/field.
The unlimited 4K bandwidth is not... unlimited.
I get the AWS is terrible, charges too much etc. But after being burnt, repeatedly, by unlimited offerings elsewhere, at some point it's just worth paying their storage price or whatever and just know they will be very unlikely to bait and switch you on pricing.
Yes, I'm looking at your unlimited backup storage data plans deals, your "lifetime" offers, your unmetered / unlimited (but wildly oversubscribed and/or we'll give you a call if you actually spin up 20 instances and max transfer on each).