It’s so notable that I can’t help but think overall watch time has flattened but they’re desperate to continue increasing revenue. Although at some point they’re likely to go into a downward spiral where they annoy too many users. I’m not there yet, but I’m on the trajectory.

Thus leaving us all permanently annoyed by the number of ads, but not quite enough to leave.

I bought it right when it came out just to not have ads on YouTube. Then when you got free streaming music included, it was impossible to turn down.

The plus side is that YouTube music recommendations seem to be quite good.

I share YouTube Premium and Apple Music family subscriptions but if pressed would give up YT Premium before I gave up Apple Music.

Also, you're comparing different products in terms of price. YT can offer a "pro" version at an additional premium (though they probably won't... Since they focus on the main use cases)

For me, it’s because YouTube is clearly going the way of cable TV, and when they finally start showing ads on those fancy paid subscriptions, I’ll have to deal with the hassle of unsubscribing.

I’ll just watch YouTube until it becomes unwatchable, at which point I’ll stop watching.

And as usual with Google, they either dont know what they are doing or they dont seems to care about YouTube Music.

What ads?

Sometimes I want to watch long podcasts about war and other times I want 30 second puppy videos and I don’t understand why I can’t have algorithms for both silos without separate accounts.

TikTok is capitalizing on this too - going to be much easier to digest a ton of tiktok then a ton of long meandering youtube videos.

Piss a user off once and you may have a much harder chance of bringing them back. Sure, you can use the fact you drove away user X in your model for how much to annoy user Y, but that doesn't un-piss-off user X. Sooner or later, there might be a significant number of users X!

For instance, I've specifically avoided "real TV" cable packages even when I can get them for cheaper than streaming TV because my use case (sports playoffs and such) is on and off, and Comcast et al have successfully trained me that it's extremely unpleasant to cancel wired cable TV, so I'd rather pay more for something less obnoxious.

It's easy to be both "data driven" and foolishly short-sighted.

As a bonus, in theory the creators get a cut, though I have no idea what it is. My clickthrough rate after years of watching YouTube is 0% so this seems like an improvement for them.

The like/subscribe/become a Patreon is also annoying, but not as bad as actual ads.

I have a zero-tolerance ad policy. If there isn't a way for me to reliably skip ads I will stop consuming the content. I'm willing to pay but all too often you end up paying for an ad-free experience and eventually ads get re-introduced anyway or the ads take on another form.

Pardon the language but fuck advertisements. If it isn't a dry, boring information-based ad I don't care for it. I have no tolerance for the emotional manipulation and psychological warfare committed by modern advertisements. I'm sick of seeing [attractive people enjoying time spent together while <using/eating/drinking> <product> during a <meal/activity/social gathering>] or [commercial designed to make you laugh or smile so you associate being happy at some level with <product>]. I was sick of those commercials 20 years ago because they're all the same and I'm even more sick of them now that I'm old enough to see how they're designed to be manipulative.

I hear you, though i can understand video makers' point of view; they are often at mercy of YT algorithm with no heads up, no recourse. Also quality content is being slowly edge out by attention grabbing mental pollution videos, actively promoted by said YT algorithm. That is one semi stable source of income for them that wont be pull from underneath them.

So I accept those ads as necessary evil. And I support few creators via patron.

Per person? No way.

At some point, everyone who is willing to give money to avoid ads will have done so and then prices will rise and thoughts of an ad-supported cheaper tier will occur. We are seeing this with Netflix right now.

What needs to change is the unsustainable capitalist model that demands ever higher growth and profits.

Also, Google is at the heart of it, paying for yt won't make one iota of change in how they think about ads and data extraction.

I realize I could do this now and it's probably better for the creators overall, but knowing that everyone I watch gets a little something, whether I subscribe to them or not, is nice.

Effectively a nerfed version of ublock origin.

Nevermind the manifest v3 changes being introduced very soon which dramatically limits the capabilities of ublock origin to some limited sub-100k static filters.

For reference I have a block list of 800k filters + dynamic filtering on Firefox.

If you want to have a sane version of the web, switch to Firefox.

It might be misguided but it’s how I feel.

Before I would click a video and check downvotes and comments for issues.

With downvotes gone, I stopped opening videos at all. I just don't trust anymore the platform to not waste my time.

Stuff like, how to change X in your Y car and then immediately find that the title is misleading. Or solutions to general IT problems. I just don't trust the platform anymore without being able to see red flags.

Short term this just means using something other than the TV app. But long term (5-15 years -_- ) I think content is going to be hosted places other than YouTube. Maybe wishful thinking, but right now if I had a YouTube link and a PeerTube link to the same content I'd send my friends to PeerTube instance.

I only need so many ads for "current SAAS desperation".

People go AWS because it’s the default, easy to hire people who know it, AWS are willing to throw credits at you if you’re in the US and once you’re on the platform it’s a serious effort to get off due to weird lock-ins.

Azure has its niche in the windows market, a lot of smaller orgs are going AzureAD as a replacement (or in addition to) a normal on-prem AD solution, once you’re on the platform it’s easy to just use more of it and consume existing contracts.

It helps immeasurably that if you’re running windows workloads they “happen” to be reasonably priced, as opposed to other cloud providers.

It’s a shame, because google cloud is definitely my favourite of the public clouds, I hope it doesn’t go anywhere and I’d be extremely happy to go back to it after using AWS for a while now.

* AWS makes the best 0th impression (the whole "nobody gets fired for choosing AWS" and as you said you can justify it based on absolutely no knowledge of it's details just purely on reputation)

* Google makes the best 1st impression, by the virtue of it's beautiful consistency of behaviour, SDK standardization, clean billing models, and overall cross-platform cohesion. It just seems like you can learn GCP faster at first.

* But AWS makes the best 2nd and subsequent impressions, as you realize consistency and clarity lacks the depth of AWS, or the true productivity gains by native services - native services Google has mostly rejected in favour of doubling down on it's K8S offering as a differentiator. While GCP can seem faster to learn, in practice building on AWS is actually faster and you can go build new full applications from scratch that you're happy operating in production quicker.

- They both throw credits at companies. - Services are similar. - Prices are similar, although Google sometimes has silly costs for "enabling" a service/API. - GCP dashboard is snappier. - AWS has good support.

I'd wager the quality of support is the main differentiator.

But I’m in Sweden and AWS keeps sending “sales engineers” who don’t know what the product is doing under the hood.

The support team I deal with has always pulled in the actual devs from AWS if we have a particular question about a service. In terms of understanding our product we send them diagrams and they are well informed.

I have a lot of information about how the services are functionally working underneath (under NDA, unfortunately) and what their limits are, they even informed us of their recommended course of action based on what we're trying to accomplish, they even invited me to feedback on their roadmap to understand what features are most desirable and even what UX patterns made the most sense to me, personally.

Of course, we were paying for enterprise support, but even during pre-sales they were giving us this information, I.E: when we were onboarding and before we had a contract in place.

We had an enterprise support contract with AWS too, that was the blessed provider at Ubisoft, but all I ever saw from them, upon multiple requests for technical assistance: was a dude telling me S3 is essentially magic, and that there's no such thing as a capacity limit in their cloud. Given that I managed to hit capacity limits in Googles and Microsofts Clouds pretty easily, I'm not sure I believe them.

I'm going to guess you're in North America? it feels like people in NA get a good support experience with AWS; however it's my experience that Europe gets absolutely shafted.

I believe the EU is following what america does in this regard, perhaps because of some potentially misplaced thought that the America is doing best practice. (or because Americans lead the world in software, so all software tends to support AWS natively).

He came back to visit us in Sweden 3 times because we kept asking for a technical contact to explain things for us.

I work for a US company now that's on AWS, the experience is very different from that.

Also, azure is the most pure play from them all and constitute the majority of MS revenue , while AWS and GCP are both a side business to the core business.

You are looking at Other Expenses/Gains. The bottom line net income was 16,436 vs 17,930 LY -- so still a Net Profit, but down 8% YoY

I don't think their equity gains/losses have anything to do with future business outlook, so ignoring this $6 billion loss YOY I'd say this was a fantastic quarter for them. They could show more in some of their non search sectors, but excluding that equity number went fro. $13 billion in profit to $17.5 for a 35% gain.

I thought buybacks were supposed to show a sign of trust in the company and that they believe the stock is "cheap"?

Anyhow buybacks are a way to give money back to investors while making each remaining share more valuable.

Buybacks are a valid tactic in corporate finance.

Isn't it the opposite? It means Google thinks their stock will be undervalued, and they want to buy it back?

It turns out there are fewer ads on their backup pages so I tend to watch there instead on YouTube.

I don't see this in the statement?

As of right now, shares are down $120, after being down $80 during the trading session so it's pretty huge.

If you haven't analyzed the revenue and expense, this number doesn't say anything. Thay may willingly be spending more to grow, but could change stance and be profitable on Monday if they wanted to. These are numbers that sound like output, but my guess is someone had a "revenue+$1B" budget, and they spent it.

I have no clue if that's the case, but it would surprise me if Google Cloud is inately unprofitable.

Good for shareholders?

GCP should be a concern for the. More than a decade in and it's still facing losses. But cloud is a long game.

(Obviously you can buy and sell smaller units but aside from that)

I have heard this exact thing debated since the late 80's, so I'm not sure what the last two years has to do with it. The fact is, whether there's a real effect there or not depends less on the precise math of the situation and a whole lot of behavioral and psychological factors. Splits goosing the stock price is a real thing that happens. It may be entirely self-perpetuating crowd-mind BS but it does still happen.

Stock split results in much cheaper options, which many retail day traders seem to prefer.

I might argue that a stock split to a cheaper price would increase your volatility due to the above. Though TSLA's current price might disagree with me as I think it's one of the most traded & is by no means cheap for options trading.

> Obviously there has never been a correlation there and even less so now with the wide availability of fractional shares.

Splits have historically caused price bumps, so I think this isn't true.

I guess all I want to say is that it's more nuanced than 'obvious' and is probably an active debate right now for economists, but I wouldn't know more exact details.

The Dow itself is a bit of gimmick in that its value is the sum of all the prices of the stocks in the index. Meaning if Google joins at $120 but kicks out a stock that was $300 then the Dow would go down 180 points.

I do wonder, with the relatively recent availability of fractional share ownership, if the effect of splits will be more muted in the future as retail is no longer limited to whole share purchases. But who knows, there's also a psychological aspect to this that's hard to account for.

Another reason that I recall is that (presumably) if a company does a split, it indicates that management is confident in the future--and a reverse split presumably indicates the opposite.

>I do wonder, with the relatively recent availability of fractional share ownership, if the effect of splits will be more muted in the future

Certainly, we've seen share prices of a fair number of companies recently that would have been considered well out of "normal" ranges historically.

Anecdotally their Google Workspace offering however almost seems to be the default for smaller business that are not IT oriented or have a separate IT department.

We went with Azure because it was the sort of “obvious” direction for a non-tech enterprise organisation that was already heavily windowsy and using office365, but what Microsoft and Amazon sell to enterprises is support. AWS didn’t start out so well here in Europe, which is what gave Azure the chance to catch up, but once Amazon caught on, they became more compliant with European legislation than Microsoft is. Things like guaranteeing that every employee that comes near your data is an European citizen, is one area where AWS is still better than Azure, but basically what they do that Google doesn’t, is that they listen and adopt to the needs of the trillion dollar industry.

When Teams rolled out as a new feature in Enterprise 365, as available to everyone, our techies called Seattle (I can’t spell redmund) and a few days later it was possible to not have it auto-available for everyone. When we had issues with Google education, and Google education was actually Google’s best attempt at being supportive of enterprise, we had to talk with a chatbot and eventually had to physically drive to Google Denmark to annoy someone to get real support because our sales rep was on vacation.

I’m not sure if you know this, but a major part of selling IT to enterprise is to sell the CTOs the ability to say “yes, email is down, but our guys are working directly with the people at Amazon/Microsoft headquarters and they are calling us with updates every 30 minutes.” or “yes our servers are in Ireland, but some of our best techies have gone and physically inspected them with the EY consultants and our DPO firm and it was completely GDPR compliant”, and all the other stuff you get to do when you make the companies billions.

In reality almost all of Google is hosted on Borg. But I worked briefly on a project that was heavily interlinked between the two. I believe there's been more efforts to move things into GCP.

G runs lots of small things on cloud but a company that is willing to fund 0.1% global efficiency efforts is never going to sacrifice far larger efficiency regressions just for the aesthetic pleasure of dogfooding cloud for major services like websearch.

I believe there had been multiple waves of attempting to get people to move to cloud, but at the time, there was no coherent strategy to move large parts of the tech stack to cloud. Unfortunately with any sufficiently complex production service that runs in borg, untangling the web of dependencies is usually too hard.

Other things work great in the cloud; if I was a Google Researcher working on public data and publishing papers, I'd intentionally use Google Cloud (with access to VMs, CPUs, and GPUs, but not any internal data like the web page crawls) with Jax, Keras, TF, etc.