Scientists plan huge European AI hub to compete with US

https://www.theguardian.com/science/2018/apr/23/scientists-plan-huge-european-ai-hub-to-compete-with-us

by charlysl • 8 years ago

213 190 8 years ago

theguardian.com

candiodari4• 8 years ago

If you had any idea what scientists make in the EU, you'd see

1) why this is happening

2) why this effort won't change a thing

They could, you know, pay competitive wages. Ridiculous suggestion, I know, at this point that'd be a 5x and more (500% and more) raise.

sveme3• 8 years ago

Places like EMBL or ESO are comparatively well paid, much better than at universities. Still less than in industry, but that's the case everywhere.

dmichulke2• 8 years ago

I fully agree.

German media constantly complains about "IT expert shortage" yet salaries grow like 3% per year. So IMO it's either not a shortage or the German industry doesn't know how to attract good people.

To put some data points there (in EUR)

1. Germany Uni Doctorate Position: 32k, 22k after deductions

2. Germany AI / Data Scientist Position: 80k, 50k after deductions

3. Luxembourg Uni Doctorate Position: 32k, 26k after deductions

4. Luxembourg AI / Data Scientist Position: 80k, 60k after deductions

Source: Friends and family and me

The above two countries belong arguably to the economical top tier, yet they can hardly compete with a US Senior Software developer (>= 100k$?)

Now imagine how the non-top tier countries fare (i.e., 80% of the EU population)

I was an AI researcher and now work as ML/AI/Data Scientist consultant (mostly for non-European companies) and I yet have to see that European job offer of 120k€ per year, hell, make that 100k€.

Radim3• 8 years ago

Another datapoint for perspective, from the Czech Republic (also an EU country):

* PhD stipend: €3.3k / year (untaxed)

* median CZ salary: €13k / year, €11k after taxes

* data scientist: €30k / year, €22k after taxes (royalty! although still below a doctorate student in Germany, apparently)

ben_w8• 8 years ago

I’m in the UK, intending to move out for obvious reasons (would’ve gone a year ago but relative got an Alzheimer’s diagnosis early 2017). Considering Berlin, Silicon Valley, or a PhD just about anywhere if the cost is in my range.

Once healthcare (at NHS level) and rent are accounted for, $100k in Silicon Valley isn’t as good as £37k in the UK. Germany has more expensive healthcare than the UK’s NHS, but significantly lower rent and overall is only slightly more expensive than the UK… unless you intend to eventually buy a house, in which case it’s suddenly much cheaper.

hdkrgr• 8 years ago

While I don't disagree with your argument, the disparity is not quite that high (talking about Germany).

A Uni Doctorate Position in Germany is 44k in the first year, ~49k from the second year (EUR before deductions) as per official TVL13 tariff. You are probably referring to the fact that it's common for PhD students to have 'half' or '3/4' positions, but afaik in CS, and especially AI, full positions are the norm.

The Industry figures seem on point.

bobcostas55• 8 years ago

I think this framing is a bit misleading. Salaries aren't exogenous, they track productivity. If a data scientist in Germany is making half of what he would be making in SV, it's because he's much less productive in Germany. Changing this isn't simply a matter of paying more to attract better talent, it's a much deeper issue.

nopinsight3• 8 years ago

What are some major reasons?

Is it because of the culture of not valuing software as much as 'hard' engineering or just lower average GDP per capita? (I read someone commenting that pay for software engineers in Japan and Korea are similarly uncompetitive for the former reason.) Swiss pay is supposed to be quite competitive with American, and other Western European countries are right around their borders.

ItsMe0000012• 8 years ago

Is it because of the culture of not valuing software as much as 'hard' engineering

No - they are paid about equally.

Here is a table (German text, does not matter, the table is clear enough): https://www.ingenieur.de/karriere/gehalt/wie-viel-verdienen-...

"Engineers with >2 years experience by sector", from top to bottom: chemistry/pharmaceuticals, automobile, energy, IT, mech. engineering, electronics, engineering and planning offices, construction

As you can see IT actually is right in the middle.

peoplewindow4• 8 years ago

European countries and companies are less likely to go big or go home in my experience. The USA has the expression, "bet the farm" and American leaders like to use it frequently. I can't think of any European equivalent saying off the top of my head. That results in bigger bets, bigger payouts when they succeed, more capital to invest in even bigger bets, etc.

The UK case of DeepMind is a good example. A big bet by British standards - lots of money raised and spent on pure research with no business model to drive it. By British standards that's incredibly brave and/or reckless. Yet a successful gamble, sold for half a billion dollars. But, sold to an American company literally called Alpha bet.

logicchains• 8 years ago

I read someone commenting that pay for software engineers in Japan and Korea are similarly uncompetitive for the former reason.

Pay for software engineers in China is actually more competitive than in Japan/Korea, in spite of China's GDP per capita being much lower. I suspect a key difference is the presence of many startups and large tech companies, something that Japan and Korea seem to lack.

hocuspocus• 8 years ago

Note, research and scientific staff is equally underpaid, if not more, in the US.

Since fundamental physics and bioinformatics have been mentioned several times in this thread, if you're curious, look into salaries at places like the National Laboratories, or even independent labs like the Allen Institute.

The market distortion in the field of AI/ML isn't as simple as poor wages in academic research. There are many strong European industries that could benefit from more AI/ML and that can definitely afford attractive salaries. Those jobs already exist around the German car makers for instance. Unfortunately, there's always more money to be made selling ads.

JumpCrisscross• 8 years ago

This reminds me of the European attempt at competing with the American BRAIN Initiative [1]:

"The key difference between Europe's HBP and the U.S.'s BRAIN Initiative is that the latter does not depend on a single scientific vision. Instead many teams will compete for grants and lead innovation into different, unplanned directions. Competition is happening via the nimh's traditional peer-review process, which prevents the conflicts of interest that plagued decision making at the HBP. Peer review is not perfect—it tends to favor known scientific paradigms—and American science funding has plenty of problems of its own. But the BRAIN Initiative's more competitive and transparent decision making is far removed from the political black box in Brussels that produced the HBP."

[1] https://www.scientificamerican.com/article/why-the-human-bra...

jamesblonde3• 8 years ago

There was a vision - it was Henry Markham's vision for simulating a cortical column. However, it was a bad idea, and it all blew up, due to a bad problem to focus on.

https://www.nature.com/news/human-brain-project-votes-for-le...

gisely• 8 years ago

The BRAIN initiative came after the Human Brain Project so it would be more accurate to call the BRAIN initiative America's attempt to compete with the European HBP.

DanielleMolloy3• 8 years ago

The BRAIN initiative started after the proposal of the Human Brain Project. The Human Brain Project also followed the Blue Brain Project, which was started in 2005. Leaving criticism of the HBP aside - it would be wrong to put it as if the HBP is the "European answer" to the BRAIN initiative.

tehlike2• 8 years ago

Unless it is supported by capitalist intent, it cannot compete with us.

ageofwant2• 8 years ago

You have already lost. Because 'us' almost certainly does not include you.

tehlike2• 8 years ago

Not sure i get you. Us=the us, and it does include me.

thaumaturgy• 8 years ago

Question for people knowledgeable on the topic: are there many examples of planned efforts like this actually working, where some group is able to attract top talent in some industry and keep it for the long term? The article mentions CERN; are there others?

I would guess that most concentrations of talent are an unpredictable side-effect of some environmental factor.

mjfl9• 8 years ago

I don't even think CERN, Higgs boson included, is worth the man hours or cash invested...

HBP doesn't seem worth it. ITER is a huge failure that is scheduled to receive billions more euros for more than 25 more years. I don't know... Europe seems to fall victim to these sort of things very readily.

riffraff• 8 years ago

ITER started as an equal partnership between EU, US, USSR (yep, not Russia) and Japan. It is insanely expensive, but I wouldn't consider it a Europe-specific madness.

randomsearch• 8 years ago

Not worth the WWW? (Amongst many other things)

stctgion• 8 years ago

Are you defining "worth it" as whether there is a profit to be made at some point? How much would you have paid for the higgs? Are some pieces of science best left in the dark because they cost too much?

throwaway77384• 8 years ago

While everyone in these comments seems to try to defend these great European projects on their merit alone, I'll try a different approach.

In my opinion, these projects are one of the last great human efforts to inspire generations with things that are not just about making more profit. But it's not about this project or that. It's about far more.

You say this money should go to hungry people instead. Let's say it's not just starving children in Africa, but maybe even a project to feed the homeless or help under-privileged children get an education...even in that case, do you actually think that if all that money was freed up, this would actually happen?

Of course it wouldn't. It would go to line even more pockets of some faceless international organisation / corporation / hedge fund / military investment buddies of some regime.

There is no way in hell that any poor or under-privileged person would benefit from the abandonment of these projects. No chance whatsoever. What little money there is, is all concentrated in the hands of a few people who have absolutely no other mission in life other than accumulating more money and power.

They have made most science happen for private sector profit-reaping already. This is some of the last stuff done for all mankind. We'd be making a huge mistake abandoning it.

Again, the people in charge will not help the poor or homeless. If you wanted to do that, you should tax multinational corporations and get rid of tax havens.

There is a kind of money hidden away in tax loopholes all over the globe that would make the LHC, HBP, heck, the ISS look like chump change.

Look there first.

sveme3• 8 years ago

I'm familiar with the European Molecular Biology Lab [1] (worked there), which is one of the best research centres in cell and molecular biology as well as bioinformatics in the world. The only real prerequisite for working there is excellence. It seems like this new AI proposal is modeled after the EMBL model, with one headquarter and several other locations around Europe. At least in the case of the EMBL, it worked incredibly well and created a great joint European atmosphere which really motivated everyone to do their best.

Obviously, there is currently much more money in AI in the industry, so I guess it will be hard for such a research centre to compete with industry. The wage differential between biotech industry and EMBL is much less than that of a hypothetical EAIL and the software industry.

[1] https://www.embl.org/

laingc2• 8 years ago

I don't disagree with your sentiment, but I think you've missed the question that GP was asking.

are there many examples of planned efforts like this actually working, where some group is able to attract top talent in some industry and keep it for the long term?

If we're talking about EMBL in Heidelberg, we can say with confidence that (along with DKFZ) it's a huge drawcard for the University. But the GP asked if it had attracted talent in some industry and kept it for the long term. Can we really say that the Rhein-Neckar region has really attracted big private sector biotech investment, on par with a place like Basel? I don't think so. By a pure coincidence of geography, companies like Roche have a presence in the general area (Mannheim I believe), but I don't think you can argue that any deliberate actions by EMBL or DKFZ have led to deep private sector talent pools.

I actually think you could turn this argument around pretty easily and show that even enormous investments in institutions like EMBL and DKFZ don't actually create the private sector spinoffs that central planners would imagine.

Like the GP, I suspect that there is a combination of environmental factors at work, that we can't fully identify at present.

carlmr3• 8 years ago

are there many examples of planned efforts like this actually working, where some group is able to attract top talent in some industry and keep it for the long term?

Many, I'm not sure, but how about Silicon Valley? Which was borne mostly out of government backing of military projects and the university system. https://en.wikipedia.org/wiki/Silicon_Valley#History_(1971_a...

ItsMe0000012• 8 years ago

Answering in a wider context:

Silicon Valley: https://youtu.be/ZTC_RxWN_xo

It wasn't planned as such, but the outcome is a result of unlimited debt-based US government spending in WWII to win the electronics war. Only after creating very fertile soil did private enterprise enter the picture.

Also, anything war related, ever, including rockets, airplanes, anything that shoots and explodes, submarines, the Internet,... Part of it is that the public is perfectly willing to have "central planning" and pretty much unlimited tax-based spending for military purposes, but try to do get them to accept the same for civilian purposes...

You may also check out the economic history of Japan and South Korea, but also how European countries like the UK rose to power. LOTS of government planning.

Then there is the entire exploration of the world, done mostly by governments, and even when it was private entities it was with very heavy government support and involvement.

Did you know that Gregor Mendel, the monk who did the experiment with the peas that is seen as the basis for genetics, was not an underemployed monk working on his own, but part of a very organized business-government-church effort to support local industry? I think it was about how to get better wool producing sheep: https://youtu.be/D8m-ZEr9qV8?t=276 (MIT Prof. Eric S. Lander)

The "market" is not magic that does everything. It only works well under specific circumstances - and those are exactly the scenarios the books will pick. Read (economist) authors like Ha-Joon Chang (https://en.wikipedia.org/wiki/Ha-Joon_Chang).

ghaff4• 8 years ago

You may also check out the economic history of Japan and South Korea, but also how European countries like the UK rose to power. LOTS of government planning.

Yes, although in the case of Japan (MITI and the like), there are counterexamples like the Fifth Generation project. In addition, there were economists like Lester Thurow in the US arguing strongly at the time that the US had to follow Japan's lead in industrial planning to create cosortia like SEMATECH, which largely didn't pan out.

T-A3• 8 years ago

I find the comparison with CERN (made by the Guardian article; CERN is not mentioned in the open letter) misleading and somewhat troubling.

As of 2016 (last year for which there are figures; new ones should be out in a couple of weeks) CERN had 82 research physicists on staff, out of a total headcount of 2531 [1].

The vast majority of researchers working at CERN are associates (982), users and visitors from other institutions, mainly universities in member countries. CERN's primary role is to provide shared facilities (accelerators and detectors) which are too expensive for individual members, not to employ researchers.

The equivalent in machine learning would be shared computing, data collection & labeling resources capable of competing with what Google, Facebook etc have to offer. A state-led initiative on shared computing resources might be more cost-effective than buying time on Google Cloud, but the main thing could well turn out to be pooling data from national databases. Yay?

[1] https://cds.cern.ch/collection/CERN%20Annual%20Personnel%20S...

nunya213• 8 years ago

OpenAI, Allen Institute for AI, MIRI, etc. are quite successful in the AI industry. Those are all in the US and pay competitive salaries though. It seems to me that Salary is a primary driver here, why stay at Cambridge for 40k gbp/yr when you could do the same research at Google Brain for $200k+ but if those two numbers aren't so out of wack I think it becomes a much more difficult choice.

jamesblonde3• 8 years ago

Our big problem in Europe is not basic research, it's systems research. We have a European platform for Big Data and AI - Hops. It's open-source. It has nice properties, but it is harder to bootstrap an ecosystem outside the valley:

(1) HDFS compatible FS with distributed metadata

(2) GPU support in HopsYARN

(3) UI to develop Keras/TensorFlow apps in jupyter, deploy on tensorflow serving. Spark support, too.

However, getting it 'out there' is still a challenge from Europe. Here is a talk on the platform at CERN last week:

https://indico.cern.ch/event/716743/

More talks here:

http://www.logicalclocks.com/eventscustom/

chrisseaton2• 8 years ago

Our big problem in Europe is not basic research, it's systems research.

Hmm maybe we mean different things by 'systems research', but I'm in systems research and we're relatively strong in this area in Europe! Notable systems groups at Cambridge, Kent, Oxford, Glasgow, Manchester, EPFL, ETH, Linz, lots of smaller ones. Major influential systems projects started in Europe like Graal. Conferences which attract Americans to fly over like ECOOP and Curry On.

jamesblonde3• 8 years ago

I agree - i would not include ECOOP and Curry On in systems research. They are prog language conferences. Systems research are the USENIX confs (NSDI, ATC, Fast), SOCC, SOSP, OSDI, Eurosys. Typically 10-30% of papers are european (often with American co-authors).

I think things have gotten worse, if anything, since Peter Druschel's article on the state of systems research in Europe a decade ago: https://hal.inria.fr/inria-00444781/document

In any case, the largest systems software company in Europe Nokia, went belly up due to a crap OS. Now we're left with SAP, who are just trying to protect last generations' software (a DBMS and its stack). Where are Europe's next big players? And why do they sell out so quick (MySQL, Elastic, JBoss, etc)?

nopinsight3• 8 years ago

Except ETH and EPFL, other names you mention do not seem to be very competitive in CS systems research with top departments in the US and even some in Asia. I am not in systems so please let us know if the ranking based on conference publications skews American or it is a fair reflection of reality.

http://csrankings.org/#/fromyear/2012/toyear/2018/index?arch...

peoplewindow4• 8 years ago

That's absolutely true but if you read between the lines it's possible that's not what he means. Lots of people in Europe use the word "Europe" when they actually mean "EU". This is such a common verbal tic it largely passes unnoticed.

The systems groups you name are mostly in the UK (soon to be ex-EU) or Switzerland (not EU, doesn't want to join). Only Linz is in an EU country with no signs of that changing.

Substituting Europe for EU and the original statement might well be correct from a purely geopolitical perspective (obviously tons of research is done in the USA by people from Europe so we're only discussing geography here and not ethnicity).

tensor• 8 years ago

US hardly has a monopoly on AI. Canada and China are also leading countries. It's great that the EU wants to invest too, the more the better! Science doesn't hold allegiance to any country and never should.

Al-Khwarizmi2• 8 years ago

The open letter does talk about North America in general (not only the US) and China. The Guardian summarized all that to "the US" in their headline.

tormeh2• 8 years ago

Doesn't this happen to US universities as well? I bet Mercedes et al could do the same as Google if they didn't require you to move to company towns in the middle of nowhere.

carlmr3• 8 years ago

I think you're thinking of Audi (Ingolstadt, Neckarsulm) or Volkswagen (Wolfsburg) middle of nowhere towns with a severe lack of estrogen. Mercedes/Daimler and Porsche are in Stuttgart, a major city (for Germany at least). BMW is in Munich, doesn't get much more major city than that.

The bigger problem I see is German company's unwillingness to pay engineers more for performance. You basically get the same salary whether you're pulling the whole team or slacking off.

tormeh2• 8 years ago

Stuttgart is not a world-class city and Munich is filled with Bavarians (I.e. conservative, burgeois). Cologne is too small, Hamburg has no tech scene. Germany only really has Berlin for the kind of tech and city that I imagine the HN audience is interested in.

Munich, doesn't get much more major city than that

San Francisco, Shenzhen, Paris, London, Berlin? Because the competition is international. Regional/National thinking is part of the problem.

ageofwant2• 8 years ago

Meanwhile, some universities had been hit so hard that they > had lost an entire generation of talented young researchers.

The schadenfreude is strong with this one. Maybe next time try treating your researchers less like cattle and start paying them market value salaries.

Radim3• 8 years ago

Trouble is, it's not really a "market". EU universities are (mostly) subsidized from other people's taxes, so calling for "market salaries" is essentially calling for increased taxation. A tough sell, especially given the distant, lukewarm connection between the people paying and the people being paid.

dosycorp• 8 years ago

Ahem ....to compete with China.

Systemic33• 8 years ago

I believe that a contributing factor to the pay gap in Europe is the fact that we have so many languages, that we are too segregated to create the same level of competitiveness as seen in the US. This means that each country is fairly protective of its people, and even though EU has free movement of labour, the language and culture barrier is strong, and internationals have a harder time to get hired, unless the speak the local language.

So the end results is that talent is spread across, and no single country has ability to become the place to be.

Joeri• 8 years ago

It also means that despite being a single market you can’t realistically launch a product across the EU. A SV startup can immediately address the whole US (if they don’t have a physical presence), but an EU startup has to go country by country. That means growth is slower, and talent is less competed for.

After brexit I’d like to see business & gov across the EU standardize on english as lingua franca to enable faster growth.

Brakenshire• 8 years ago

After brexit I’d like to see business & gov across the EU standardize on english as lingua franca to enable faster growth.

I wonder whether this will become easier now when the UK leaves, because the English language won't be tied to the competition between the major states.

This is definitely a shame for London, which I think was emerging as the centre of English-speaking Europe, I expect that role now will be more spread across a number of northern European cities, like Stockholm, Copenhagen, Amsterdam etc.

saiya-jin• 8 years ago

National egos will probably prevent that. I simply can't imagine French people speaking anything but french in their home country. Anytime I visit the place (which is almost every weekend), I see how much they don't like speaking english, although their level is OK and they are perfectly understandable. Italy might be the same, Spain too.

Northern europe has much stronger usage of english and level is much higher (I recall old grannies responding in perfect english around Oslo, and that was almost 20 years ago).

madeofpalk• 8 years ago

As someone from Australia, I actually take the opposite impression from that. These startups are faced to deal with being international by day one, making scaling out to more and more countries easier.

For US startups, on the other hand, the US is a large enough market that they might get ‘stuck’ there and need to build up more ‘inertia’ to scale out internationally

simias5• 8 years ago

After brexit I’d like to see business & gov across the EU standardize on english as lingua franca to enable faster growth.

Do you imply that brexit might make it easier for that to happen? That sounds counter-intuitive to me, once the UK leaves I believe that the only country in the EU who has English as an official language will be Ireland and even there it's cohabiting with Irish Gaelic.

Beyond that I've always been torn on this issue, on one hand having a lingua franca across the EU would be amazing and English is probably closest to achieving that, on the other hand English is effectively the language of American imperialism and its cultural hegemony. In an ideal world I'd prefer something less politically loaded and tied to a foreign superpower, like Spanish, Swedish or Romanian for instance (I'm picking random European languages who cannot be suspected of having any kind of cultural overreach in present day Europe).

Of course in this case maybe practicality trumps ideology and we should just accept our English speaking overlords for the sake of convenience.

erispoe• 8 years ago

Depends, in regulated industries you might have to adapt to product state by state in the US much more that you would have in the EU, where a lot or regulations are now more uniform.

bitL5• 8 years ago

lingua franca

Hmm, I seem to recall we already had such a language in Europe since middle ages, it started with L and ended with n. It's still dominant in medicine. It could be funny to use it in tech EU-wide as well with "proper German pronunciation" ;-)

rorykoehler• 8 years ago

Definitely agree re English being adopted as the lingua franca (that really sounds weird haha). Been saying this for years however there are too many fragile egos in European politics that feel that adopting English is an attack on their identity. There was even talk of removing English from EU legal texts altogether after brexit

acct17712• 8 years ago

Is this not preferable to one country "buying all the best players" and exerting their will on others?

peoplewindow4• 8 years ago

Is that really true? Everyone in science speaks English and DeepMind, based in the UK, is in fact one of the world's top AI research hubs. Their papers are consistently excellent and of course DeepMind's workforce comes from around the world.

That's the European AI hub, right there.

The issue here isn't the language barrier. The issue is one that goes largely unstated in the ELLIS letter (it's a call for action, not really a plan). Which is this ... remind me what's wrong with working with the Americans again?

DeepMind was founded by a Brit - the guy who once worked on Theme Park and other much loved British games. He built up an absolutely top class research team out of nowhere, based purely on VC funding, then sold to Google not only because of money but because Google also built a world class AI research team out in California, and it made sense to join forces so they could work together on things like TPUs.

And why shouldn't the European AI hub have joined forces with the US AI hub? What benefit could there be from preserving institutional and funding barriers between Google and DeepMind? The letter sort of dances around the edges of this question, but ultimately the answer is purely politics.

It says things like

"This weakens Europe"

and

"we want the best basic research to be performed in Europe, to enable Europe to shape how machine learning and modern AI change the world"

and

"many European companies whose future business crucially depends on AI are not perceived as competitive"

These are all political goals, worse, they are not logical thinking at all - not a good look for a bunch of scientists.

AI research is entirely open. Having a new academic research hub in the EU-that-isn't-the-UK won't change the access those European countries have to the underlying research papers or techniques. They can use AI just the same regardless of where the research is being done. Maybe China would try to close up their researchers, but America and the UK certainly won't.

So if European companies are "perceived as uncompetitive" or are not "shaping the world" due to their poor use of AI (this is a dubious theory to begin with) then the solution is for more European companies to download TensorFlow and Torch and get cracking. It isn't for money to be poured into academics who will take it, write a few papers and then go straight into industry anyway.

T-A3• 8 years ago

remind me what's wrong with working with the Americans again?

Make that American companies and the problem, from an academic perspective, is that they recruit away students and researchers from academia. That works for a few years, but longer term, the argument goes, it depletes academia to the point that new talent can no longer be trained... by academia.

At that point, the companies wishing to keep recruiting will create their own training facilities and/or step up collaboration with academic institutions. That's where the American part comes in: if the existing talent has all gone to the US and those making the decisions have been there all along, they are likely to focus their training efforts on the US.

On top of that, you have the well-publicized problems with data collection by foreign entities and local tax avoidance, which are indeed political.

gaius4• 8 years ago

These are all political goals, worse, they are not logical thinking at all - not a good look for a bunch of scientists

Indeed, it’s almost on the level of Make Europe Great Again

kome• 8 years ago

No, language doesn't matter: EU has lower wages because it tax (and redistribute) more, and I like it this way.

enriquto3• 8 years ago

and even though EU has free movement of labour, the language and culture barrier is strong, and internationals have a harder time to get hired, unless they speak the local language.

Certainly not in all fields. In all the math labs that I know (from France, UK, Italy) more than half of the researchers are from other countries.

Of course, if you want to teach, it is much easier to get hired if you already speak the language.

(edit: format of the quote)

KKKKkkkk1• 8 years ago

A smaller labor supply implies that the supply and demand curves cross higher. Pay should be higher then.

kitd• 8 years ago

Though the segregated markets could also imply smaller demand.

segmondy• 8 years ago

From an outsider, I can tell you what. It's not the langauges. It's the culture, it's the work/life balance. It's things like working reasonable hours, spending time with family & friends. Having long holidays, take time off from work to care for a newborn. It's things like GDPR.

America is brutal, America is a gladiator pit, winner take all. Money above all things, the religion is capitalism. Make money or die trying and roll over anyone in the way.

chrshawkes• 8 years ago

I'm not sure that's the case because the United States is also filled with people that speak different languages.

gaius4• 8 years ago

the scientists say the proposed Ellis institute is essential to avoid brain drain to big tech firms

They don’t say tho’ the mechanism by which it will accomplish this. What’s in it for the individual researchers who choose this path over an industry job? ITER only exists because there is no well-funded private-sector fusion research. If Musk or Bezos or Gates entered that game it and CERN would evaporate overnight.

I’m no AI guru but I’m a pretty decent infra engineer who can wrangle storage and compute at scale, and they’ll need people like that too, why would someone like me choose this path? What’s the offer?

Genuinely curious.

molteanu• 8 years ago

Well, a nice office in a "culturally rich city with dozen of events, museums and world-class cuisine"?! Like it's done for all other industries. Some nice polished marketing and the fear of missing out if you don't go to the place where all the "exciting things happen".

gaius4• 8 years ago

Well, a nice office in a "culturally rich city with dozen of events, museums and world-class cuisine"?!

I have been thinking about this today, and I'd take a cut in pay of say 25% in exchange for a private office in which to do interesting work, and absolute job security followed by a final salary pension. The way it used to be.

But that actual deal would be a pay cut to 25%, crammed into a office too dingy for Civil Servants, at risk of being TUPE'd at the drop of a hat. No thanks.

icc973• 8 years ago

Where this is shown up starkly is in Self-driving startups. This list of startups [0] only has 1 or 2 in Europe out of 33.

Mercedes does most of it's self-driving research in the US.

[0] https://angel.co/job-collections/top-self-driving-cars-start...

DoctorOetker• 8 years ago

is it just me or does the guardian symbol in the browser tab (white G on a black disc) always look like the pirate party symbol (black flag on white background in a black circle)?

l5870uoo9y• 8 years ago

The question is really if Europe can compete with the whole US ecosystem without protectionist policies (and there justified). Nowhere is the US domination more visible than in the tech sector, where Europe essentially is a tech vassal state. The European countries need to build not only an AI hubs, but also an competitive business ecosystem that attracts top talent through salaries, opportunities, challenges and benefits. I don't think there is much appetite in the European population for competing on lavish salaries and raw capitalistic incentiments (both categories in which US leads).

European leaders can just look at China and see successful protectionist policies. It makes sense to have some policies not only in US-EU relationship, but also inside EU that spreads the talent on a regional level so regions don't just get drained. This is also a big issue for East- and Central-european countries.

adventured2• 8 years ago

There's actually something fascinating about the dominance / vassal premise: it applies within the US, and it applies within Europe as well.

Silicon Valley, Boston, NY, Seattle, and a few others dominate the US on tech, with the rest of the US mostly fitting into your tech vassal state premise.

The exact same thing is true within Europe for industry and tech. Germany overwhelmingly dominates Europe on auto manufacturing for example. What kind of cars are Russia, Spain, Portugal, UK, Poland, Ukraine, Greece, etc producing? The Germans are out selling and out producing the rest of Europe to an epic degree.

How many Airbus planes - and or what share of components - are manufactured in Romania, Czech, Greece or Poland? In Europe these are industrial vassal states to France or Germany.

dragandj• 8 years ago

The headquarters are in Germany, but tiny Slovakia, for example, has huge car production facilities.

ocschwar2• 8 years ago

I don;t know about AI, but when it comes to using formal methods to nail down software development and make it reliable, the EU is already outpacing the US.

ra1n85• 8 years ago

This is an assertion without data to back it up. Care to elaborate?

ocschwar2• 8 years ago

INRIA, where Coq development takes place.

aje403• 8 years ago

Considering the vast majority of all CS and modern software languages and practices have originated in the US in the first place and the US still, by any rational or unbiased measure, has the largest and most innovative tech companies in the world, I am not totally convinced of what you're claiming

alexhutcheson2• 8 years ago

The headline seems misleading. They're not planning one "huge hub" in one location. Instead, they're planning a bunch of mid-sized institutes spread across different countries, each with "hundreds of computer engineers, mathematicians and other scientists".

That approach is understandable politically, but it seems much less likely to have the intended effect. If the goal is to kick-start a knowledge cluster in Europe that can compete with the Bay Area, then it seems much more likely to succeed if the funding and talent is concentrated in a single site. "Hundreds of computer engineers" is equivalent to a mid-size remote office of a major tech company.

That's just an observation, though. I realize that the choice probably isn't this plan vs. one mega-site. More realistically it's this plan vs. nothing at all, or vs. competing piecemeal efforts by various EU member states.

bitL5• 8 years ago

I think it would go standard European way - build an institution by a top-down directive reflecting political will, shake each others hands, then a few years later cry that you have all this outdated infrastructure but no inventive people working for you nor any bleeding-edge results, as all relevant people are snatched up and overpaid by industry giants.

Eridrus• 8 years ago

I think a hundred researchers/engineers is well beyond a midsize company remote office, because the caliber of people and mission will be very different. DeepMind is less than a thousand people, and was probably on the order of a few hundred when it became famous.

alexhutcheson2• 8 years ago

For comparison, the Fraunhofer Society has 72 institutes in Germany, all focused on applied research. Collectively, these institutes employ ~25,000 scientists and engineers[1], which gives us an average of ~350 scientists and engineers per office. This seems very comparable to what's being proposed, both in mission and in scale of individual institute.

This seems to be an effective model for applied R&D, but I haven't seen anyone argue that a Fraunhofer Institute is sufficient to be a seed for a major tech cluster. The scale just isn't big enough.

[1] https://www.fraunhofer.de/en/about-fraunhofer/profile-struct...

Barrin92• 8 years ago

I think it's worth experimenting with a more diverse and distributed model because the 'one megahub' structure also has disadvantages, low spread beyond the borders of the tech industry being one example.

Especially countries like Germany with diverse industrial production spread out across the country might benefit more from having technology clusters spread out as well.

Even Silicon Valley companies like Google have at least partially taken on this approach by opening labs in Toronto or Paris, Amazon plans to create about a hundred jobs around Stuttgart and Thübingen.

throw2016• 8 years ago

SV has always been heavily subsidized. A lot of the fundamental tech comes from government research projects. Startups have access to easy capital and there is no transparency of the 10/1 or 100/1 failures, how this is sustained and where all the money is coming from.

There is definitely govt money including from the CIA, NSA at work here. So there is no reason the EU cannot do the same, infact they are late.

The open source movement has also played a key role in most of the SV success stories in the recent past. Zuckerberg, Page and others are the celebrated 'free market' 'wealth creators' but the work of Stallman, Torvalds, Rasmus and thousands of others have undeniably added billions of dollars of value.

Our currently ideological narrative of 'wealth creators' and 'free markets' completely fails to account and value these and other critical social investments. We make the narratives we want.

All the SV money further subsidizes companies like Uber and other startups hungry for market share, who can under price and run out competition in other markets which is dumping. Its just that no one is going after them in Europe and elsewhere yet but the rules are there and can be enforced when needed.

youpassbutter• 8 years ago

This is great news. We need more tech hubs and more competition in the tech industry. I'm still confused as to why europe decided to cede the entire tech industry to the US. Do europeans love us this much? That they want us to have all money/technology?

But this doesn't go far enough. Europe has 600 million people. They should have 2 or 3 tech hubs.

Look at what china is doing. They are developing their own tech industry. That is what europe should do. Foster their own tech industry.

giardini• 8 years ago

I want to thank everyone who contributed to this thread. The variety of experiences and viewpoints provided an incredible insight to someone who's been mostly in the USA for the last 2 decades. My conclusions:

a) There's not a snowball's chance in Hell that a European AI hub compete with the US,

b) Europe remains far more fragmented than I thought it was; even more fragmented than it appeared when I was in high school. "Brexit" may actually increase Europe's (and Britain's) organization.

c) Britain and Europe have tons of good people but a lack of organization suitable for pure research, be it by governments, corporations, "hubs", or non-governmental organizations.

d) China is similarly SNAFU'd but worse in some respects (unsettled markets, lack of transparency & property rights). Luckily, there doesn't appear to be a ghost of a chance of having one of China's many languages adopted as a scientific "lingua franca". [Note to self: cross the (formerly) relevant item off my "bucket list".]

e) Perhaps restoring near-universal use of the character encoding UTF-16 or even US-ASCII is not a completely dead idea but merely one to be delayed until the adoption of English as the "lingua franca" of the Internet and Europe, salvaging much-needed bandwidth. Time will tell.

Scientists plan huge European AI hub to compete with US

Related Stories