Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Open Euro LLM: Open LLMs for Transparent AI in Europe

openeurollm.eu

Post reply on HN
Print view View original

181–190 of 289 posts

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#181

Post by lyu07282 »

jansan wrote: post #104
bayindirh wrote: post #67

Earlier quoted context omitted.

I can see that you're unfamiliar with how EU grants and how these project collections work, but I don't have much time to address this with great detail. As a person who's in this type of projects for a long time, what I can say is "it works", because people do not compete with each other, but will build it together. What I can say is, if they have came this far, there's already plans about what to do, and how to do,…

In the case of Quaero [1] "it did not work". Sure, all involved parties were praising the project and by constantly shifting goalposts they could label it a success, but in the end it was a huge waste of money, sucked in by the usual suspects. [1] https://en.wikipedia.org/wiki/Quaero

> it was a huge waste of money, sucked in by the usual suspects

other example: https://en.wikipedia.org/wiki/Gaia-X

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#182

Post by varjag »

mft_ wrote: post #177
varjag wrote: post #171

Earlier quoted context omitted.

If your metric of innovation is the amount of rockets exploded at debuts you shouldn't bring up SpaceX really. The EU had committed to a number of deep space and scientific instrument programmes spanning decades and seen them through to success. It operates its own GNSS constellation. It is second only to NASA. Calling it a failure is ridiculous.

I suspect the previous poster's metric of innovation was more along the lines of: * developing the first meaningful fully-reusable first stage rocket, and continuing to develop it to the extent that no other launch systems are even in the same ballpark as regards cost, cadence, or mass to orbit * developing, and continuing to develop, the only full-flow staged combustion rocket engine * developing, and continuing to…

The snarky point of explosions wasn't mine, the poster I replied to brought it up.

The rest of your points is really one item, launch vehicles. It's where the USA clearly has the lead (above everyone else, not just ESA in particular). The question was whether the EU can successfully manage complex projects and it clearly can, suggesting otherwise is delirious.

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#183

Post by Maken »

menaerus wrote: post #90
bayindirh wrote: post #84

Earlier quoted context omitted.

This is a Horizon grant, meaning: - There's continuous reporting, and money is not guaranteed. - You can be removed from the project by not meeting project goals on time. - (In this case) There are corporations which are planning to commercialize this thing. - There's a concrete and sound roadmap, and it's evaluated in a competition by an independent body and got selected. - Without a sound landscape survey, you can'…

I am familiar with Horizon and EU grant funds and how these projects are managed - basically a big waste of time and money without concrete results with plenty of bureaucracy crawling the project ideas and resources to the halt. I am not saying that because I think people are not hard working, I am just saying that the process as it is is not fit to produce something (market) competitive. Process is there to serve it…

>Can you list me few examples of successful projects you have in mind? Why, for example, Mistral AI isn't among them?

Yeah, I was surprised too. I would expect Mistral to jump on the band wagon if only to get some easy EU funding.

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#184

Post by varjag »

closewith wrote: post #174
varjag wrote: post #155

Earlier quoted context omitted.

The Large Hadron Collider clearly works and is infinitely more complex than yet another language model.

The LHC is comparably complex to LLMs and CERN is a cautionary tale. The web was invented there and yet the vast majority of the subsequent innovation and economic benefit coalesced in the US because the European countries were culturally unable to capitalise on it.

LLMs are not "comparably complex" to LHC, they are perhaps 0.1% of the complexity from engineering POV. By any metric, from system design to physical infrastructure to run them.

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#185

Post by Lionga »

CrimsonRain wrote: post #113
tobyhinloopen wrote: post #97

Better late than never. I can't wait to get my hands on a mediocre AI that's 2 generations behind!

2 is generous. 5 is more likely. Also can't wait to get bombarded with cookie popup, ai bias popup, then ai accuracy popup etc.

Anything more then a powerpoint coming out of this would be a generous expectation.

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#186

Post by closewith »

throwaway473825 wrote: post #180
closewith wrote: post #173

Earlier quoted context omitted.

> It operates its own GNSS constellation. Only 33 years later and mostly launched on Russian rockets, behind GLONASS and BeiDou. > It is second only to NASA. Calling it a failure is ridiculous. In what respect? Space? Certainly not, far behind the US and Russia and questionably competitive with China. Economically? Behind US and China. R&D? Behind US and China. Manufacturing? Behind US and China. You are refusing to…

Not all satellite systems are equal: https://www.bloomberg.com/news/articles/2023-09-20/russia-s-... >The US constellation isn’t as accurate as the newer networks, said Roberts, the Sydney-based professor. “It used to be GPS was out in front,” he said. Now, though, the EU’s Galileo is in the lead, with China’s BeiDou close behind, he said. It's basically Galileo > BeiDou > GPS >>> GLONASS.

That would be expected for a system launched 33 years later, but in Galileo and GPS are identical for civilian use (and obviously no-one knows the military capabilities of Block III satellites as that's undisclosed).

GPS+Gailleo is the current SOTA, but it's nonsense to say Galileo is "best".

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#187

Post by Xelbair »

varjag wrote: post #152
misiek08 wrote: post #99

Earlier quoted context omitted.

"It works!" is the only thing that will be visible on web page after hundreds of milions will be burned. I’m observing few of such „unprecedented” cooperation projects from EU funds. A lot of meetings, a lot of managers, plenty of very unskilled people creating mess and few names doing presentations so companies will believe everybody know what are they doing. Same from company side - they need being in those project…

Europe runs its space programme in this way and so far it has pretty good track record. There are more ways to build stuff than the worship of personality.

Europe honestly had pretty... mid track record for space program.

should i bring up Galileo and how many years it took?

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#188

Post by Maken »

bayindirh wrote: post #67
isoprophlex wrote: post #57

I'm HIGHLY sceptical. The academics will love it, because they get money. But look at that list of parties involved. More than twenty parties supplying people; none of them will have this initiative on the top of their list of loyalties and priorities. Meaning, everyone will talk, noone will take charge, some millions change hands and we continue with business as usual. Instead this should have been a single new non-…

I can see that you're unfamiliar with how EU grants and how these project collections work, but I don't have much time to address this with great detail. As a person who's in this type of projects for a long time, what I can say is "it works", because people do not compete with each other, but will build it together. What I can say is, if they have came this far, there's already plans about what to do, and how to do,…

While I do think EU grants are a good thing, I'm sceptic about these too-big-to-fail multi-national projects. I still remember the Human Brain Project.

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#189

Post by closewith »

varjag wrote: post #179
closewith wrote: post #173

Earlier quoted context omitted.

> It operates its own GNSS constellation. Only 33 years later and mostly launched on Russian rockets, behind GLONASS and BeiDou. > It is second only to NASA. Calling it a failure is ridiculous. In what respect? Space? Certainly not, far behind the US and Russia and questionably competitive with China. Economically? Behind US and China. R&D? Behind US and China. Manufacturing? Behind US and China. You are refusing to…

Russia had its last deep space mission (failed) in 1996. GLONASS did not operate until 2005. Chinese contributions to scientific space missions had been very modest although am sure they may catch up later. NASA has 3x the budget of ESA. The question was if the EU method of doing project works and it does in very unambiguous manner.

> GLONASS did not operate until 2005.

Five years before Galileo.

> The question was if the EU method of doing project works and it does in very unambiguous manner.

As the EU falls economically and scientifically behinds what used to be our peers, it's obvious that it _doesn't_ work. Refusing to recognise that reality is a spectacular example of the Ostrich effect,

Re: Open Euro LLM: Open LLMs for Transparent AI in Europe

#190

Post by Xelbair »

throwaway473825 wrote: post #180
closewith wrote: post #173

Earlier quoted context omitted.

> It operates its own GNSS constellation. Only 33 years later and mostly launched on Russian rockets, behind GLONASS and BeiDou. > It is second only to NASA. Calling it a failure is ridiculous. In what respect? Space? Certainly not, far behind the US and Russia and questionably competitive with China. Economically? Behind US and China. R&D? Behind US and China. Manufacturing? Behind US and China. You are refusing to…

Not all satellite systems are equal: https://www.bloomberg.com/news/articles/2023-09-20/russia-s-... >The US constellation isn’t as accurate as the newer networks, said Roberts, the Sydney-based professor. “It used to be GPS was out in front,” he said. Now, though, the EU’s Galileo is in the lead, with China’s BeiDou close behind, he said. It's basically Galileo > BeiDou > GPS >>> GLONASS.

And frankly accuracy does not matter.

for navigation using Code method GPS-tier is basically good enough.

for precise measurement you use phase measurement of the signal, and what you care about is good(low) DoP of constellation and amount of satellites within sight-line - not from which system they come(to oversimplfy it a bit)

Post reply on HN