Hacker Newsnew | past | comments | ask | show | jobs | submit | Nitrolo's commentslogin

Any particular template you'd recommend? My resume is LaTeX too but I'm not 100% happy with it (about 98% happy and much happier than with anything else however).

I can't find the one I used now. But I just searched "latex resume template" and picked one that I liked. Some good ones at https://www.overleaf.com/gallery/tagged/cv

For years YouTube would randomly change my language to Italian and auto translate all the titles, which was annoying but I can change the language back to English and it was fine.

Thing is, I haven't lived in Italy for well over a decade. I don't consume much Italian content. I've tried my best to purge my Google account of any possible trace of Italian language settings, I've never saved an Italian address etc.

And yet Google must be convinced that, because I was in Italy over a dozen years ago when I made the account I must be Italian, and that cannot possibly change.


That's already how this works, basically. You have to keep paying a fee to keep your patent, and the fees increase with each year of the patent.

The fees are fixed, but the effect is pretty similar to what you describe.

In some jurisdictions you get a discount on the fees if you declare that you are open to licence out your patent. Then everyone can use your invention without having to ask for permission,as long as they pay the license fee.

EDIT: I looked it up, in Germany (§ 23 PatG) the patent office can set the licence fees that others have to pay, so you can't play games by just setting the fee to a trillion euros.


> effect is pretty similar to what you describe.

The max yearly fee is $7,400? It can be a lot or absolutely nothing depending on a specific patent.

> In some jurisdictions you get a discount on the fees if you declare that you are open to license out your patent.

Would it really ever be worth the effort just to save a few thousand per year at most?

3,700 if you're a non profit or < 500 employees.


§23 PatG covers the case where the patent owner wants to use a general public licensing scheme offered by the patent office to avoid having to deal with every single licensee. That one is opt-in by the patent owner.

§24 PatG offers a compulsory licensing scheme, but it requires not only that the would-be licensee made an effort to find an agreement with the licensor, but also that such a compulsory license is in the public interest.

Since the public interest doesn't seem well-defined in PatG, you'd have to check older cases to see how that turns out, but I'm not sure if even "doing so reduces bandwidth use on the internet" is enough: you still have the option of using another codec and compensate the worse compression by going for lower resolution video, for example. 4K video is not a human right ;-)

(and "we hog internet resources" is the last thing Netflix et al want to say out loud anyway: ISPs are eager enough as-is to try to get them to pay up for access to their customers.)


The difference would be that the fee you have to pay is scaled to the value someone else can get from the patent. So if you have a patent which is not valuable to you but is valuable to someone else, then you either have to sell the patent to them or pay a tax based on what value they could get from it.


Why are you not allowed to hire a cook, painter or child carer?


Not allowed to use pre-tax money for it. You can do it with post-tax money just fine.


Can't deduct the expenses.


Just the obligatory note this isn’t deductible in the US either.


Child care is deductible in Switzerland, but maybe its also due o the fact its so horribly expensive.


Child care is deductible in the US too, for the vast majority of people.


Does anyone know of a nice UI wrapper for something like whisper.cpp?

I need to write a lot of long texts for work and some good dictation software would be great. I know there's Dragon, but somehow I have not been able to find something that fits my need and is free.



Do you know if these implementations also support leveraging the M1/M2 GPU, such as shown here? https://github.com/openai/whisper/pull/382


Google Recorder transcription is very good: https://play.google.com/store/apps/details?id=com.google.and...




+1 pon MacWhisper -- I pay for the pro version because I find it to be just great. Sindre's Aiko is excellent too!



Lawyer in training here!

Trademarks do get grouped into classes (the so called "Nice Classification", like the French city [0]). The classes are very detailed, and the first 34 refer to "Goods", the rest to "Services".

Specifically class 9 refers to (among many other things), "computers, computer software", class 42 to "design and development of computer hardware and software".

However those are mostly just to simplify searching for and dealing with the huge amount of trademarks. Each trademark comes with its own list of goods and services it's supposed to cover. Their trademark in the UK [1] covers "computer software, software and apparatus for the extraction of business information and knowledge".

They also appear to have registered a trademark this september [2] with a much broader scope, so after Meta started using threads. I'm not familiar with UK law and how it relates to trademarks in this case, however.

[0]: https://en.wikipedia.org/wiki/International_(Nice)_Classific... [1]: https://trademarks.ipo.gov.uk/ipo-tmcase/page/Results/1/UK00... [2]: https://trademarks.ipo.gov.uk/ipo-tmcase/page/Results/1/UK00...


Quite a bit, but Wikipedia provides sources! So I checked [0].

In section 3.3 it says that GSK argued their patent on the new form was being infringed because "there were batches of anhydrate that converted almost entirely into hemihydrate when stored at 40 °C and 75 % humidity within one month."

So "if you keep it wet and warm it will partially convert over the course of a month", and as far as I can tell the "seeding" aspect was not clarified, at least not in the article.

Also it bugs me that the Wikipedia author decided to call the loss for GSK a "technicality". It was decided that it was not relevant that microscopic quantities of the new form were created by accident, I think that is the correct decision and not a technicality.

[0]: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4479028/


Is there something like oobabooga to easily run this in a click-and-run way? Where I can load up a model, a text, and ask it questions?


See my comment here: https://news.ycombinator.com/item?id=38020655 for a CLI tool that lets you do this.

Note that embedding models are a different kind of thing from a Large Language Model, so it's not the kind of model you can ask questions.

It's a model which can take text and turn it into an array of floating point numbers, which you can then use to implement things like semantic search and related documents.

More on that here: https://simonwillison.net/2023/Oct/23/embeddings/


The Hugging Face page for the model has a two line load-and-encode Python code demo: https://huggingface.co/jinaai/jina-embeddings-v2-base-en


iirc ooba has its own integrated vectordb called superbooga.

I bet you could hack this in.


As a comparison, an estimate for the cost of World War II is $5T, for US spending at least [0].

Obviously there's differences in methodology etc. but seems to be the same ballpark. Not sure I'd call WWII cheap.

[0]: https://www.stlouisfed.org/on-the-economy/2020/february/war-...


Depending in the demographics, a fertility rate below replacement level can still lead to a rising population even when excluding migrants, as the people who would be "replaced" by the new births might live for many more decades.

I.e., we'd probably still have a housing crisis in Europe even if not a single foreigner came in.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: