This actually sounds like a great idea. I procrastinate a lot. I also play a lot of Raid Shadow Legends. I should totally gate those purchases behind hitting my goals.
This is off to a great start. I've been studying German through various methods for around 3 years now. Mostly as a mental exercise but also because I love German television.
For me, reading has been one of most productive ways to learn.
Dark was terrific. Time travels and parallel worlds. Deutschland '83 (in particular) but also '86 and '89 were quite decent - about a young German becoming a Stasi superagent of sorts.
Oh, and while you're at it, get Das Boot if you haven't already, also available as a TV miniseries.
As for movies, you could do a lot worse than picking up a few Wim Wenders movies - say, Der Himmel über Berlin or Alice in den Städten.
Oh, if you pick up Der Himmel über Berlin, consider picking up In weiter Ferne, so nah! as well. Same universe, but more run-of-the-mill than Himmel über Berlin, which is probably the most poetically beautiful movie I have ever seen.
(All non-amazon links will probably require a German IP / VPN - you may find those shows on other streaming platforms, too)
Stromberg - German adaptation of The Office, but Bernd Stromberg (aka Micheal Scott / David Brent) is a bit of an asshole this time and it's not a paper company, but an insurance company.
Tatort - mostly serious show about police solving murders, produced by all of the local public broadcasters, e.g. there are Tatorts from Cologne, Münster, Munich, each using the regional dialect. "Polizeiruf 110" is the version of the ex-GDR, it works the same way. The Münster-Tatort is quite funny:
Tatortreiniger - a show about a crime scene cleaner in weird situations. If you know "The Cleaner" from the BBC: This is the original show it's based upon. The main character uses a northern german / Hamburg dialect.
Hubert und Staller - show about incompetent bavarian rural police solving murders, heavy focus on comedy, bavarian dialect. The first few seasons are quite funny, the later ones are meh.
I liked: Türkisch für Anfänger, Dark, Kleo, ans Stromberg. Unfortunately I’d say IMHO if you’re accustomed to US series you won’t find a lot of German ones that stand up to the production values and acting quality. Dark is a notable exception there I think.
Yes, but "Dark" (Netflix) doesn't have a lot of words spoken per hour, as I recall, so perhaps it's not so good for language learning. There's also "Beat" (Amazon) with a somewhat gruesome main storyline.
If you wanted to get linguistically flummoxed there's "Capitani" on Netflix. (A great story, but it's in Luxembourgish!)
I'm originally from Tennessee. Chapel Hill actually. Boyhood home of Nathan Bedford Forrest. There's a monument to the guy and our streets were literally named after him. I've seen firsthand the systemic racism that permeates the miseducated people of that State. That said, this doesn't surprise me.
I'm just glad I move to Austin and away from that crap.
This is exactly what I'm working on! My project is taking Zoom conversation, using pyannote for speaker diarisation, whisper for transcription, pinecone.io for semantic search, then feeding that into GPT-3 so we can ask questions about conversation.
For us this is super useful because it's not unusual for our discover sessions to last days and we're all terrible at taking notes.
As a nerd, my brain is already buzzing on ways that I could use this for my groups D&D campaigns.
Are you getting good results when summarizing a human speaking? On my project, even though Whisper does a good job translating it, I'm not happy with the query results. My theory is that GPT-3 is designed for written word and the way people speak and the way they write are structurally different. Or I'm just figuring this out and I'm not good enough at it yet.
It’s often not enough to just index the snippets themselves. You may need to augment them. For instance, you may need to keep track of the context, and prepend it to the actual snippet that you want to index.
The important thing in such a pipeline is not GPT 3. The important thing is the retrieving/ranking algorithm that finds the most relevant snippets and feeds them into GPT 3. The latter is only the mouthpiece, if you will.
In fact, you might even find that you’re better off without it (no confabulation, ground truth data).
I've got tons of notes so it shouldn't be too hard to do a write up. Currently it's in a private repo, but if I can get sign-off from my boss I'll open source it.
that's like saying VHS vs betamax is subjective. VHS won but wasn't better. Or Minidiscs vs CD's. It's mostly due to marketing fails and people hopping onto marketing bandwagons and pre-estalbished products.
TS is more popular due to combination of google dropping the ball in defining the standards and getting others onboard with them cause they overestimated the leverage they have, failure at marketing it and how everyone is already bought into Javascript.
"What is stored on the blockchain is most often a link to a web page[6] that itself points to the associated object. Here again, we have lost all idea of decentralization"
No where in the article did the author mention IPFS (https://ipfs.io/). The trend with NFTs is that the digital assets are stored on IPFS, typically through a pinning service like Pinata (https://www.pinata.cloud/). With the asset being content addressable and not location based, and the address being a permanent part of the blockchain, how can the above statement be true?
It is very rare I wouldn't call it a "trend" yet. Also, I agree that using IPFS links addresses the problem of content integrity, but the link can still rot, and having to pay a centralized service to pin your data (even in a decentralized network like IPFS) does not solve the decentralization and disintermediation problems.
Why these platforms don't use hash of a file inside blockchain as bare minimum. Shouldn't be too expensive to store that data? Hash would represent the "original" no matter where is it hosted on the future.
The way I understand it (and I might be totally wrong here) is that ipfs links do contain the hash of the file stored, so by putting an ifps link as the token URI you are both providing the content hash as well as a way to actually locate the content.
The surviving NFT projects that will require no active maintenance in the next 10 years will be the ones who fully dedicated to on-chain only (cryptopunks, chainrunners, etc).
Unless there is an active community that maintains its strength during market cycles (like BAYC), current NFT project death will be 95% over the next 10 years, just like a start-up.
There are also plenty of NFTs with the images embedded directly onchain. Whether base64 binary with an image mimetype or encoded as an SVG.
NFT users and buyers will begin to care as domain names for dead projects expire over the next year.
To me, it is really strange how people see flaws in an implementation and don't even look for the other people that are working on improving those flaws. The market is saying "hey, you perceive flaws, go make hundreds of millions offering improved products, you don't even need outside capital to do so!" but I guess it's always been this way, have fun writing blog posts.
I'm not sure the question? An NFT with no reliance on third party infrastructure, as described above, would have no ability for denial of service. More analogous to a non-digital art piece, where only third party individuals or companies that are digitally presenting an existing piece could be coerced by the arm of the state to stop displaying it via a DMCA request, but the existing piece still exists for anyone physically present to see. The NFT would just transcend geography in this case.
So a marketplace like OpenSea or a digital gallery could potentially receive a DMCA request. Galleries with unknown owners are likely going to become more prevalent.
It's up to the communities and potential purchasers to be discerning about what they interact with or buy, not really a standard to place on the technology thats a higher standard than other media. NFTs aren't here to solve plagiarism and unauthorized distribution. For a previous NFT work being re-released by someone else, consumers should check with the official community first before trying to acquire that contract. For an NFT work that wasn't previously issued (like a famous artist putting their stuff onchain suddenly now), they should do greater due diligence.
so basically we should throw away millennia of history and trial and error to reboot our societies using a technology that doesn't even work without an internet connection and a very fast computing device?
the DMCA is 24 years old and its current use relies on an internet connection more heavily than onchain NFTs do. what stimuli were you replying to exactly, its not clear to me.
the DMCA relies on something called a law and the power of the State to enforce such law, using the force if necessary.
your suggestion is to give up centuries of refinements on law making/enforcement and rely on personally checking every single trnsaction, because if you don't and are scammed, you're screwed and there's nothing you can do to get your money back.
This thread isn't about NFTs being an improvement over all copyrighted work, this thread is about NFTs being an improvement over other NFTs.
On the consumer side I wasn't suggesting anything, I said what they have to do in response to an already existing market with already existing transactions. This isn't about some future reality we imagine on an internet forum, this is about one you are simply not in.
If you think checking is too hard, then make a service that makes it easier to check and charge people for it, how often do you rule out business models that you personally identify, because thats what the market is saying to do.
It's a spreadsheet that is a list of tokens created by artists, and the corresponding owner in the column to the right. That spreadsheet will survive the bankruptcy of any particular company, which gives the tokens a sense of permanence, and Benjaminian aura if you will, that you don't get from a ledger run by Google.
That is what is is, that is what people find it useful for. It is not supposed to do any other things.
> It's a spreadsheet that is a list of tokens created by artists
or by people that found something slightly interesting and put it on the blockchain for money.
here's no assurance that some artist made it, there's no assurance the author is the one selling it, there's basically no assurance, apart that they are trying to trick people into thinking that it's an investment.
It's, in other words, a scam.
> That spreadsheet will survive the bankruptcy of any particular company
It won't.
Monuments in Rome have survived thousands of years and hundreds of wars.
98% of NFTs will last at max few months
> that you don't get from a ledger run by Google
at least I can sue Google if they publish on their ledger the hash of something I made and someone else sold without my permission...
> That is what is is, that is what people find it useful for
Please, don't try to sell me your snake oil.
That's the opposite of why people want them.
People absolutely don't care about the item, proof is that the majority of items sold as NFTs would never ever have a market outside of the speculative bubble (and in fact they never had one before).
Correct. I do not see how it follows that a NFT minted by the artist who made it is a scam.
> 98% of NFTs will last at max few months
You make this mistake because you think the url matters. It does not. It is the collectors job to ensure the art survives. Agreement on what the token represents, either ad-hoc or through a hash is enough.
> at least I can sue Google if they publish on their ledger the hash of something I made and someone else sold without my permission...
True.
> People absolutely don't care about the item, proof is that the majority of items sold as NFTs would never ever have a market outside of the speculative bubble (and in fact they never had one before).
There is a speculative bubble, but art will remain hard to sell the less it is a short-term speculative investment, but the creation of this previously non-existent market is indeed the contribution here. Without the market, the spreadsheet is not interesting.
Thanks for sharing.