Idk, the assertion that there are not tech giants in Europe because Europeans are simply too independent or whatever seems a bit far fetched. For one, there are - Spotify for instance, and Spotify...
Idk, the assertion that there are not tech giants in Europe because Europeans are simply too independent or whatever seems a bit far fetched. For one, there are - Spotify for instance, and Spotify also does many of the things that people complain about US tech companies, from going into deals with Joe Rogan to "enshittification" as much as the latter has lost all meaning.
Or Klarna, the posterboy for irresponsible financial decisions. All European.
I would not use mastodon as a success story, but outside of that nitpicking (throwing in curl as a replacement) I think a lot of the reasoning holds true. It is good that we're not building tech...
I would not use mastodon as a success story, but outside of that nitpicking (throwing in curl as a replacement) I think a lot of the reasoning holds true. It is good that we're not building tech giants in that way. I think we are lagging behind in some important ways. E.g. for cloud hosting there are no alternatives inside eu that are even close to aws or gcp when it comes to features and maturity.
Maybe it's my own naivety, however I don't see this as a negative. There are plenty of cloud hosts in Europe; they're just not massive monopolies. I'm assuming based off my own experience that...
I think we are lagging behind in some important ways. E.g. for cloud hosting there are no alternatives inside eu that are even close to aws or gcp when it comes to features and maturity.
Maybe it's my own naivety, however I don't see this as a negative. There are plenty of cloud hosts in Europe; they're just not massive monopolies. I'm assuming based off my own experience that most companies don't need hyperscaling, they need a beefy server and someone to manage it.
I'm of the mindset that every business and their kitchen sink moving to "the cloud" was a mistake because it's tightened the grip of a few massive tech monopolies that will never be displaced.
I'm not talking about a lack of cloud hosts, I'm talking about the services they offer. I also think it's a problem that the services are not standardized making it hard to move from one provider...
I'm not talking about a lack of cloud hosts, I'm talking about the services they offer. I also think it's a problem that the services are not standardized making it hard to move from one provider to another, I think that should reasonably be caught by antitrust laws. But it's a different problem from the services being not being offered in eu at all.
Out of interest, what kind of services would you miss if you had to use Hetzner or Scaleway rather than AWS or GCP? I’ve always kind of written off any of the vendor-specific services because of...
Out of interest, what kind of services would you miss if you had to use Hetzner or Scaleway rather than AWS or GCP?
I’ve always kind of written off any of the vendor-specific services because of those exact interoperability/lock-in problems you mention - maybe it’s a case of me not missing what I never had, but to me they’ve always seemed proprietary for the sake of being proprietary. I don’t see a huge amount of innovation in those services (but again, that could be unfamiliarity or general dislike colouring my view, so I’m genuinely interested in how people use them!), they tend to strike me as being mostly driven by revenue targets at the provider side.
Things like postgres, docker, terraform, Linux itself - and yeah, the conceptual idea of cloud services allowing the hardware to be rented dynamically by the minute or hour rather than by the month or bought outright - all provide a ton of value to me in the “allowing me to do things I otherwise couldn’t” sense, but hyperscalers thankfully don’t have a monopoly on any of those.
Just to chime in here as I work with both GCP and Hetzner, and Hetzner is very much DIY compared to all the managed services you can get with GCP. Hetzner gives you basic tools to orchestrate a...
Just to chime in here as I work with both GCP and Hetzner, and Hetzner is very much DIY compared to all the managed services you can get with GCP. Hetzner gives you basic tools to orchestrate a network and VMs, but the rest is up to you. With GCP we can basically just do a few clicks, then we have a fully managed Kubernetes setup, a database with backup and HA. Stuff like good log search is just there for all your stuff, whereas in Hetzner we have to deploy our own Loki for example. Of course Google has quite a pricetag on those services, but still way cheaper than the man hours needed to provide something similar in Hetzner. Also, in the last five years, GCP have only had one outage affecting us when their HTTP load balancer when down globally. Hetzner frequently has odd smaller outages, weird issues and is generally more flaky.
This took some time for me to figure out, sorry about that, but it's more of a feeling than anything. At work I've used mostly GCP and some AWS, privately I'm investigating setting up something...
This took some time for me to figure out, sorry about that, but it's more of a feeling than anything. At work I've used mostly GCP and some AWS, privately I'm investigating setting up something for myself and so far I've looked mostly at Scaleway and Hetzner, and a little bit at OVHCloud. So goals are obviously completely different. Honestly it's more of the overwhelming list of services offered by AWS (and to some extent GCP) compared to the listings of the european alternatives. The services I've investigated for myself can probably be well covered inside EU from what I've seen, and I didn't spend any time actually looking for alternatives for the services we use at work, mabye it can be replicated.
From what it looks like so far, Scaleway is the most competent. Also I know that I didn't talk about that in the original post, but Scaleway only have presence in 3 countries from what I can see (except edge nodes). That part alone would be complete no go for at least parts of what my company wants to deploy.
So, to summarize, there's no specific thing I know that I would be missing, and maybe all of the extra things in AWS list of services is crap. Hope that made sense =)
Please keep the government's regulatory fingers out of my tech stack. This would be terrible for innovation.
I also think it's a problem that the services are not standardized making it hard to move from one provider to another, I think that should reasonably be caught by antitrust laws.
Please keep the government's regulatory fingers out of my tech stack. This would be terrible for innovation.
As a non-European (rather, a dirty American), I feel like this is a little bit of cope. Like yeah, that's great that Europe and Europeans contributed greatly to the foundations of the modern...
As a non-European (rather, a dirty American), I feel like this is a little bit of cope. Like yeah, that's great that Europe and Europeans contributed greatly to the foundations of the modern Internet. And contributions will continue to be made and will be made for a long time to come! A lot of cool tech comes out of Europe.
But like Bwerf says, where are the at-parity hyperscalers in Europe? Us Americans are weaponizing interdependence, even towards our allies and friends (maybe I need to add "former" as a qualifier, sigh), and Europe has no alternatives. OK, maybe there doesn't need to be an European alternative for TikTok or Douyin. I think there's an argument that NO ONE needs these in the first place...But how about AI? Seems like all the big players are American or Chinese. That said, I just learned about Mistral and its "Le Chat." Though I don't know how widespread its use is. Can anyone chime in?
From discussions I’ve read previously, there’s two issues: Lack of unified VC markets (making financing and scale-up harder) Lack of a common language. Probably one of the biggest issues. In the...
From discussions I’ve read previously, there’s two issues:
Lack of unified VC markets (making financing and scale-up harder)
Lack of a common language. Probably one of the biggest issues. In the U.S. you can spin up a new app that supports more than 500 million people in North America with just three languages (English, Spanish, and French). In China you can do the same for a billion people with just Mandarin (and maybe Cantonese). To do something in Europe, you need two dozen languages for an equivalent number of people.
Huh. I guess on the language front, I assumed that those who need to work with these platforms and tools know English. Because they're already using American-based tools. And so having tools in...
Huh. I guess on the language front, I assumed that those who need to work with these platforms and tools know English. Because they're already using American-based tools. And so having tools in English would be enough. Very Anglocentric of me, I know.
But I guess I hadn't considered if there are any regulatory/legal requirements by the EU or by specific countries within that require software to have local/native language options. Especially if used by governments. If laws like that do exist, I could see how having a diverse number of languages -- normally a good thing -- could be an impediment, at least in the start-up phase.
On the VC market front, that's interesting that's it not unified, even though there is a common market and currency. That said, I know very little about all that, so maybe not that unusual!
I’m looking at it more from the customer side. A U.S. company can get by with just English, Spanish, and French UIs for, say, a food delivery app. But a company based in the Netherlands catering...
I’m looking at it more from the customer side. A U.S. company can get by with just English, Spanish, and French UIs for, say, a food delivery app. But a company based in the Netherlands catering to the European market as a whole has to include Dutch, French, Spanish, Slovak, Swedish…
Now, AI-powered translation might help reduce this cost/difficulty, but there’s already the entrenchment of American firms that will make it very hard for Euro-google to get off the ground.
I'm not really convinced the language is as big of a factor as you think it is. It's a factor for a company that wants to grow its customer base in Europe, but if we're really talking about...
I'm not really convinced the language is as big of a factor as you think it is. It's a factor for a company that wants to grow its customer base in Europe, but if we're really talking about something like European Google, they'd be targeting a global customer base more or less immediately regardless of where they're based.
I don’t think language would be the problem, or at least, not big enough to be considered a barrier. I would sooner point out that we are a lot of different countries with different laws. A...
I don’t think language would be the problem, or at least, not big enough to be considered a barrier.
I would sooner point out that we are a lot of different countries with different laws. A service that is legal in Spain might not necessarily be in Portugal for example. Bureaucracies are also different.
I know that this can also happen in the US, but I have the impression that their laws are more standardized than ours.
Deepmind are British, although they were bought by Google a fair way back, so I guess they’ve become American through acquisition? Same goes for Arm - UK company that’s done very foundational...
Deepmind are British, although they were bought by Google a fair way back, so I guess they’ve become American through acquisition? Same goes for Arm - UK company that’s done very foundational technical work but is now owned by offshore capital, with ongoing acquisition attempts from US tech giants.
I agree with you about the article being a bit skewed, perhaps even bordering on smug, but I think the lack of hyperscalers is the opposite to the example I’d choose! The incentive structure for those guys is to capture as much of the market as possible, increase prices as far as they’re able, and make it as hard as they can for customers to leave - all “value creation” and success in the economic sense, but not something that I’d say we actually need or want in terms of making things better for developers or end users. I want a thriving competitive market of cross-compatible small players who actually have to compete on price or service!
I don’t have time to write out the whole essay in my head - maybe I’ll draft it as a standalone post one of these days - but I do think that this touches on something really important: the way we define value, and the way we incentivise “success”. Because what the article absolutely does get right is how much real innovation comes out of university research labs, open source developers, small scrappy organisations, and occasionally even government initiatives. But sharing and using those innovations is seen as failure from an economic perspective - or at best, hidden and unquantified success, like the wider business impact of all those billions of devices running Linux - whereas taking innovations from the original creators, locking them away, and charging what the market will bear is seen as success. Breaking compatibility to make it hard for customers to leave is seen as success. Acquiring competitors and increasing prices is seen as success.
The more time I spend in the world of science, technology, and research, the more convinced I am that large companies are a net drag on overall progress - not because they’re inherently evil, or even because they explicitly intend to stand in the way of positive developments, but because acting in a way where that’s a major side effect is the rational move from their perspective. Motivated nerds, when given sufficient resources, make actual progress driven by a combination of self-actualisation and ego - often for decent but not amazing compensation in the world of academia. Companies and investors at a small enough scale figure out how to package and distribute those innovations in a way that’s palatable to the world, which academics are often otherwise broadly terrible at. But companies at a larger scale are directly driven to buy out those smaller ones, with the explicit goal of reducing competition, increasing prices, or both.
Who’s going to stand up for that broke PhD student who invented something amazing? Who is going to stand up for works that no one else is talking about? This is what is valuable in the world. This is what pushes humanity forward. This is what makes the world better for us. But there is a problem. It doesn’t make crazy money. [...] Just think about it: from a business point of view, taking a bit longer to give you a better, more detailed video about a paper is not a virtue. It is an operational defect that needs to be solved immediately. But not around here it isn’t.
I wish more services had just noped out and left to be honest. Somewhat selfishly as I don't use most big tech (Apple being the exception), but certainly not any social media (is Tildes social...
I wish more services had just noped out and left to be honest. Somewhat selfishly as I don't use most big tech (Apple being the exception), but certainly not any social media (is Tildes social media - probably kinda).
Mistral is definitely one of the big players in the space, fwiw, and competes with the big American AI companies. Its valuation isn't as absurd as OpenAI's or Anthropic's, but it's definitely big...
Mistral is definitely one of the big players in the space, fwiw, and competes with the big American AI companies. Its valuation isn't as absurd as OpenAI's or Anthropic's, but it's definitely big enough that you can't write off Europe when it comes to big AI companies.
That honestly might be a blessing in disguise. I’m more than convinced that the bubble WILL burst, and OpenAI and Nvidia will be the epicentre of the incoming storm. Smaller AI companies might get...
That honestly might be a blessing in disguise.
I’m more than convinced that the bubble WILL burst, and OpenAI and Nvidia will be the epicentre of the incoming storm.
Smaller AI companies might get away with “minor” injuries (relatively speaking). If Mistral was as big as OpenAI and had their financial accounting as absurd as OpenAI, the french - and possibly the rest of the EU - could end up paying dearly. As I’m sure it will happen with the americans
Strictly as an anecdote: I know at least three successfull IT company from EU (that I was working with or have friends employed) that was bought by USA or Canada consortiums.
Strictly as an anecdote: I know at least three successfull IT company from EU (that I was working with or have friends employed) that was bought by USA or Canada consortiums.
The world needs a real competitor to Google Search, and I'm not aware of anything comparable being undertaken by Eurozone companies or institutions. An independent index and transparent...
The world needs a real competitor to Google Search, and I'm not aware of anything comparable being undertaken by Eurozone companies or institutions. An independent index and transparent algorithmic ranking would go a long way towards establishing trust and setting a standard that advertising is not required to support what should be a public service.
I missed this thread when it was posted but my late entry is this: Putting the author's steriotypically French delivery aside, I think their point is fantastic. What if we looked at value...
I missed this thread when it was posted but my late entry is this: Putting the author's steriotypically French delivery aside, I think their point is fantastic. What if we looked at value differently? We should do that.
What if Torvalds is more successful, as a human being, than Jeff Bezos? If the score isn't kept in dollars it's not a difficult case to make.
I know this seems naive to the point of absurd from an American perspective, where the market devours idealism without even trying, but from humanist perspective it's entirely valid.
Idk, the assertion that there are not tech giants in Europe because Europeans are simply too independent or whatever seems a bit far fetched. For one, there are - Spotify for instance, and Spotify also does many of the things that people complain about US tech companies, from going into deals with Joe Rogan to "enshittification" as much as the latter has lost all meaning.
Or Klarna, the posterboy for irresponsible financial decisions. All European.
I would not use mastodon as a success story, but outside of that nitpicking (throwing in curl as a replacement) I think a lot of the reasoning holds true. It is good that we're not building tech giants in that way. I think we are lagging behind in some important ways. E.g. for cloud hosting there are no alternatives inside eu that are even close to aws or gcp when it comes to features and maturity.
Maybe it's my own naivety, however I don't see this as a negative. There are plenty of cloud hosts in Europe; they're just not massive monopolies. I'm assuming based off my own experience that most companies don't need hyperscaling, they need a beefy server and someone to manage it.
I'm of the mindset that every business and their kitchen sink moving to "the cloud" was a mistake because it's tightened the grip of a few massive tech monopolies that will never be displaced.
I'm not talking about a lack of cloud hosts, I'm talking about the services they offer. I also think it's a problem that the services are not standardized making it hard to move from one provider to another, I think that should reasonably be caught by antitrust laws. But it's a different problem from the services being not being offered in eu at all.
Out of interest, what kind of services would you miss if you had to use Hetzner or Scaleway rather than AWS or GCP?
I’ve always kind of written off any of the vendor-specific services because of those exact interoperability/lock-in problems you mention - maybe it’s a case of me not missing what I never had, but to me they’ve always seemed proprietary for the sake of being proprietary. I don’t see a huge amount of innovation in those services (but again, that could be unfamiliarity or general dislike colouring my view, so I’m genuinely interested in how people use them!), they tend to strike me as being mostly driven by revenue targets at the provider side.
Things like postgres, docker, terraform, Linux itself - and yeah, the conceptual idea of cloud services allowing the hardware to be rented dynamically by the minute or hour rather than by the month or bought outright - all provide a ton of value to me in the “allowing me to do things I otherwise couldn’t” sense, but hyperscalers thankfully don’t have a monopoly on any of those.
Just to chime in here as I work with both GCP and Hetzner, and Hetzner is very much DIY compared to all the managed services you can get with GCP. Hetzner gives you basic tools to orchestrate a network and VMs, but the rest is up to you. With GCP we can basically just do a few clicks, then we have a fully managed Kubernetes setup, a database with backup and HA. Stuff like good log search is just there for all your stuff, whereas in Hetzner we have to deploy our own Loki for example. Of course Google has quite a pricetag on those services, but still way cheaper than the man hours needed to provide something similar in Hetzner. Also, in the last five years, GCP have only had one outage affecting us when their HTTP load balancer when down globally. Hetzner frequently has odd smaller outages, weird issues and is generally more flaky.
This took some time for me to figure out, sorry about that, but it's more of a feeling than anything. At work I've used mostly GCP and some AWS, privately I'm investigating setting up something for myself and so far I've looked mostly at Scaleway and Hetzner, and a little bit at OVHCloud. So goals are obviously completely different. Honestly it's more of the overwhelming list of services offered by AWS (and to some extent GCP) compared to the listings of the european alternatives. The services I've investigated for myself can probably be well covered inside EU from what I've seen, and I didn't spend any time actually looking for alternatives for the services we use at work, mabye it can be replicated.
From what it looks like so far, Scaleway is the most competent. Also I know that I didn't talk about that in the original post, but Scaleway only have presence in 3 countries from what I can see (except edge nodes). That part alone would be complete no go for at least parts of what my company wants to deploy.
So, to summarize, there's no specific thing I know that I would be missing, and maybe all of the extra things in AWS list of services is crap. Hope that made sense =)
Please keep the government's regulatory fingers out of my tech stack. This would be terrible for innovation.
As a non-European (rather, a dirty American), I feel like this is a little bit of cope. Like yeah, that's great that Europe and Europeans contributed greatly to the foundations of the modern Internet. And contributions will continue to be made and will be made for a long time to come! A lot of cool tech comes out of Europe.
But like Bwerf says, where are the at-parity hyperscalers in Europe? Us Americans are weaponizing interdependence, even towards our allies and friends (maybe I need to add "former" as a qualifier, sigh), and Europe has no alternatives. OK, maybe there doesn't need to be an European alternative for TikTok or Douyin. I think there's an argument that NO ONE needs these in the first place...But how about AI? Seems like all the big players are American or Chinese. That said, I just learned about Mistral and its "Le Chat." Though I don't know how widespread its use is. Can anyone chime in?
From discussions I’ve read previously, there’s two issues:
Huh. I guess on the language front, I assumed that those who need to work with these platforms and tools know English. Because they're already using American-based tools. And so having tools in English would be enough. Very Anglocentric of me, I know.
But I guess I hadn't considered if there are any regulatory/legal requirements by the EU or by specific countries within that require software to have local/native language options. Especially if used by governments. If laws like that do exist, I could see how having a diverse number of languages -- normally a good thing -- could be an impediment, at least in the start-up phase.
On the VC market front, that's interesting that's it not unified, even though there is a common market and currency. That said, I know very little about all that, so maybe not that unusual!
I’m looking at it more from the customer side. A U.S. company can get by with just English, Spanish, and French UIs for, say, a food delivery app. But a company based in the Netherlands catering to the European market as a whole has to include Dutch, French, Spanish, Slovak, Swedish…
Now, AI-powered translation might help reduce this cost/difficulty, but there’s already the entrenchment of American firms that will make it very hard for Euro-google to get off the ground.
I'm not really convinced the language is as big of a factor as you think it is. It's a factor for a company that wants to grow its customer base in Europe, but if we're really talking about something like European Google, they'd be targeting a global customer base more or less immediately regardless of where they're based.
I don’t think language would be the problem, or at least, not big enough to be considered a barrier.
I would sooner point out that we are a lot of different countries with different laws. A service that is legal in Spain might not necessarily be in Portugal for example. Bureaucracies are also different.
I know that this can also happen in the US, but I have the impression that their laws are more standardized than ours.
EU Inc. might be the solution for this though
Deepmind are British, although they were bought by Google a fair way back, so I guess they’ve become American through acquisition? Same goes for Arm - UK company that’s done very foundational technical work but is now owned by offshore capital, with ongoing acquisition attempts from US tech giants.
I agree with you about the article being a bit skewed, perhaps even bordering on smug, but I think the lack of hyperscalers is the opposite to the example I’d choose! The incentive structure for those guys is to capture as much of the market as possible, increase prices as far as they’re able, and make it as hard as they can for customers to leave - all “value creation” and success in the economic sense, but not something that I’d say we actually need or want in terms of making things better for developers or end users. I want a thriving competitive market of cross-compatible small players who actually have to compete on price or service!
I don’t have time to write out the whole essay in my head - maybe I’ll draft it as a standalone post one of these days - but I do think that this touches on something really important: the way we define value, and the way we incentivise “success”. Because what the article absolutely does get right is how much real innovation comes out of university research labs, open source developers, small scrappy organisations, and occasionally even government initiatives. But sharing and using those innovations is seen as failure from an economic perspective - or at best, hidden and unquantified success, like the wider business impact of all those billions of devices running Linux - whereas taking innovations from the original creators, locking them away, and charging what the market will bear is seen as success. Breaking compatibility to make it hard for customers to leave is seen as success. Acquiring competitors and increasing prices is seen as success.
The more time I spend in the world of science, technology, and research, the more convinced I am that large companies are a net drag on overall progress - not because they’re inherently evil, or even because they explicitly intend to stand in the way of positive developments, but because acting in a way where that’s a major side effect is the rational move from their perspective. Motivated nerds, when given sufficient resources, make actual progress driven by a combination of self-actualisation and ego - often for decent but not amazing compensation in the world of academia. Companies and investors at a small enough scale figure out how to package and distribute those innovations in a way that’s palatable to the world, which academics are often otherwise broadly terrible at. But companies at a larger scale are directly driven to buy out those smaller ones, with the explicit goal of reducing competition, increasing prices, or both.
[Edit] A quote from a video I linked a few weeks back says it better and more succinctly than I can, actually:
Mistral the cat. Sounds awesome.
Their logo is pretty clever, too. Like definitely looks like a cat peaking above a wall, but the cat in pixel art form looks like an M, for Mistral!
Gods damn you imgur, may you burn in Avernus.
https://imgur.kageurufu.net/dB165ng.png
Yes, it is a cute logo. I am happy
Ahh tbf i support imgurs stance. Gods damn you repressive government regulations!
(And thanks for the alt link, you're a hero.)
I wish more services had just noped out and left to be honest. Somewhat selfishly as I don't use most big tech (Apple being the exception), but certainly not any social media (is Tildes social media - probably kinda).
Mistral is definitely one of the big players in the space, fwiw, and competes with the big American AI companies. Its valuation isn't as absurd as OpenAI's or Anthropic's, but it's definitely big enough that you can't write off Europe when it comes to big AI companies.
That honestly might be a blessing in disguise.
I’m more than convinced that the bubble WILL burst, and OpenAI and Nvidia will be the epicentre of the incoming storm.
Smaller AI companies might get away with “minor” injuries (relatively speaking). If Mistral was as big as OpenAI and had their financial accounting as absurd as OpenAI, the french - and possibly the rest of the EU - could end up paying dearly. As I’m sure it will happen with the americans
Strictly as an anecdote: I know at least three successfull IT company from EU (that I was working with or have friends employed) that was bought by USA or Canada consortiums.
The world needs a real competitor to Google Search, and I'm not aware of anything comparable being undertaken by Eurozone companies or institutions. An independent index and transparent algorithmic ranking would go a long way towards establishing trust and setting a standard that advertising is not required to support what should be a public service.
Well, there is European search perspective, which has had a partial go-live in France.
I missed this thread when it was posted but my late entry is this: Putting the author's steriotypically French delivery aside, I think their point is fantastic. What if we looked at value differently? We should do that.
What if Torvalds is more successful, as a human being, than Jeff Bezos? If the score isn't kept in dollars it's not a difficult case to make.
I know this seems naive to the point of absurd from an American perspective, where the market devours idealism without even trying, but from humanist perspective it's entirely valid.