Well that paper only says it's theoretically not possible to completely eliminate hallucination. That doesn't mean it can be migitated and reduced to the point of insignificance. I think fabricating things is part of creativity. I mean LLMs are supposed to come up with new text. But maybe they're not really incentivised to differentiate between fact and fiction. I mean they have been trained on fictional content, too. I think the main problem is to control when to stick close to facts and when to be creative. Sure, I'd agree that we can't make them infallible. But there's probably quite some room for improvement. (And I don't really agree with the premise of the paper that it's caused solely from shortcomings in the training data. It's an inherent problem in being creative and that the world also consists of fiction and opinions and so much more than factual statements... But the training data quality and bias also has a severe effect.)
That paper is interesting. Thanks!
But I really fail to grasp the diagonal argument. Can we really choose the ground truth function f arbitrarily? Doesn't that just mean given arbitrary realities, there aren't hallucination-free LLMs in all of them? But I don't really care if there's a world where 1+1=2 and simultaneously 1+1=3 and there can't be an LLM telling the "truth" in that world... I think they need to narrow down "f". To me a reality needs to fulfill certain requirements. Like being contradiction free etc. And they'd need to prove that Cantor applies to that subset of "f".
And secondly: Why does the LLM need to decide between true and false? Can't it not just say "I don't know?" I think that'd immediately ruin their premise, too. Because they only look at LLMs who don't ever refuse and have to decide on a truth.
I think this is more related to Gödel's incompleteness theorem, which somehow isn't mentioned in the paper. I'm not a proper scientist and didn't really understand it, so I might be wrong with all of that. But it doesn't feel correct to me. And I mean the paper hasn't been cited or peer-reviewed (as of now). So it's more like just their opinion, anyways. I say (if their maths is correct) they just proved that there can't be an LLM that knows everything in any possible and impossible world. That doesn't quite apply because LLMs that don't know everything are useful, too. And we're concerned with one specific reality here that has some limitations. Like physics, objectivity or consistency.
LoL.. everyone saw this coming. And if the deal happens and the Chinese get to price gouge the Russians, they will build massive storage and suck Russia dry for pennies on the dollar.
They already started without US weaponry. They've hit a shitload of oil infrastructure in the past month, one as deep as 1400km past the border, using large fixed wing kamikaze drones
I meant it's a start for the US allowing the use of American weapons on Russian soil. I think their military and economic infrastructure should be fair game as well.
you mean the migration 'crisis' and collapse in '"living" standards' which were brought on by US-EU neoliberalism driving down the standard of living in other parts of the world before coming home to roost?
there are certainly ways of reversing direction, but people in the core would sooner choose literal fascism before giving up their imperial lifestyle. they use the IMF to politically terraform 'underdeveloped nations' and export their own harms so they can say they're 'meeting climate goals', and then complain about all the emissions and migrants coming from those countries which are ravaged to supply their hyperconsumption. the same migrants which predominantly staff their service, medical and technology sectors to prop up their precious treats and their oh-so superior 'knowledge economies'.
voting for fascism is the individualistic choice which lets them keep their treats and means they don't need to interact with their neighbours or advocate for real change. it's easier to blame the victims of their actions than to cut the DARVO shit and accept responsibility.
Showing once again that fascism is much too palatable to most "moderates". FdI is a literal offshoot of the original Fascist Party and many of its members, Meloni included, have publicly said they idolize Mussolini. Naturally they had to say they're sorry for saying it out loud, which means they definitely changed their views, right?
A key piece missing here: many EU nations were working for a long time towards reducing their reliance upon gas & oil in general and on it coming from Russia as a source specifically. iirc Germany was even on target to reduce that to zero by 2025, before the Ukraine invasion started? So how much of this dip in revenue was already likely, regardless of the war?
It would seem disingenuous to call it a "loss" even in that case, if it was less than a prior year but it wasn't expected to begin with - that's generally not how that word works, when the identical revenue stream projected into the future was not "expected".
Gazprom may even be okay with everything that is happening as part of the overall restructuring / realignment plan to shift from EU to more Asian markets where impending climate change may be considered to be less of an impediment to future purchases than the EU market was becoming. e.g. when faced with market changes, rather than pivot to meet the newer challenges of green(-er) energy, Gazprom may welcome this doubling down on old ways of doing business in the Asian sphere.
This week, the Kremlin issued a decree to halt Gazprom dividends this year. A strange move if it 'was expected to begin with', especially if we cpnsider that the Kremlin relies heavily on Gazprom to fuel its war machine.
And the alternative markets in Asia may or may not come for Russia's future, but they don't come anytime soon as Russia simply lacks the infrastructure to these destinations. According to an Analysis by the Atlantic Council, building the pipelines would cost USD 100 billion, and you can't build them overnight. That takes a very long time, time Moscow doesn't seem to have.
So call it a loss or whatever you want, but Gazprom is in big trouble, and so is the Kremlin.
There's nothing 'disingenuous' about calling it a 'loss'. It clearly states they posted a 7 bill NET loss. Not projected, not relative to the past... net. loss. They paid $7 bill to stay in business this year and are cancelling dividends. Their sales of gas reduce by more than 50% in 1 year. They aren't 'welcoming' anything. They are praying doing business in Asian markets will be enough to bring them back to profitability and the latter half of the article makes it clear this is very unlikely for a few years at least.
This is imo the best news I've read all year in that it shows strong evidence the sanctions are working. I just hope they hurt enough that Russia retreats from Ukraine, and gives up their plan of continental conquest for good, but I'm not holding my breath.
A key piece you forgot to mention is Nordstream 2, which was only completed in 2021. Prior to the war, both Russia and Germany were planning to increase natural gas delivery to Europe, not decrease it, and certainly not to eliminate it by 2025!
And you can be sure that if Gazprom is reporting a 7 billion loss, you can be sure they worked as hard as possible to make that number as small as possible. I'm willing to bet a lot of money the actual loss is a lot bigger.
Corporate real estate speculators are no longer interested in owning commercial office spaces. Does this mean RTO failed? And how is it not a win to get rid of corporate ownership so that real people can afford real estate again?
Property prices are used here as a proxy for the health of the economy.
The author may or may not be concerned about the affordability of homes. But that's a different issue. This is just about absolute prices. And low prices mean low interest in Germany. As the article says, people are interested more in London or Paris.
This leads to fewer investments in the economy and Germany is at risk of falling behind.
If office buildings are in demand in one place but not the other, what could that say about their respective economies (i.e. the companies that would use those office spaces)?
It's one indicator of economic health out of many. Unlike many other assets, real estate is immovable. So it has a strong association with the location.
Yes, work from home is a good argument, why falling office prices might not be a big deal. The question then becomes, is this trend more pronounced in Germany than in other countries? Maybe, but probably not.
Also unfortunately the article mentions Signa (commercial) and vonovia (residential) real estate. So there's no clear separation. But overall the trend is true for both.
So it seems this is mainly in reference to large commercial building rather than homes (excluding apartments).
Its interesting to read that german property owners don't like to reduce property prices in comparison to other countries when I thought others were pretty against lowering prices as well.
reuters.com
Heiß