grumbelbart2

I by no means wanted to suggest the opposite (that's how I found the connection)!

Schmidthuber is famous for his groundbreaking work in machine learning, but also for being somewhat left out when people list the "Grandparents of Deep Learning" and for being vocal about that. I just wanted to poke that bit. If your work takes off, it would be a great chance for him to shine.

Funnily enough, Jürgen Schmidthuber invented / coined the term "Gödel Maschine" in 2003: https://arxiv.org/abs/cs/0309048

He'll probably have a field day over this.

Of course it exists in Germany. If I say ‚Hello‘, I could not claim copyright on that sentence because it does not have the required schöpferische Höhe (threshold of originality). The sentence is simply not copyrightable, putting it into public domain.

Note that this is different from Gemeinfreiheit, where a copyrighted work essentially becomes public domain 70 years after the death of its creator.

Compiler output is a derivated work of the source code and thus keeps the copyright of the code.

That would violate a different law, though, § 201 StGB (German penalty code) forbids recording if there is a reasonable assumption of privacy (which you can have even in public).

This complaint is about a different law (§ 8 TDDDG), that explicitly forbids selling and owning devices that are designed as every-day objects and that can record sound and/or video and that.

No, they would not be in violation. And an independent third party can pick them up and train with them. But you are not allowed to "support" that third party, otherwise you violate the ToS.

It might depend on your jurisdiction, but LLM output is usually not copyrightable because it is not made by a human. There might be a few exceptions, if you guided the LLM in very specific and particular ways. But in general, the output of an LLM is public domain the moment it was produced.

Can you never tell anybody and just slap a license on it? Sure.

The issue here is not the license, but that you violate their ToS (if that is valid and enforceable is a different question of course). But if you publish the LLM output on github, and someone else takes it to train their LLM, and you did not actively encourage or help them, it's fine.

Nobody "owns" it for the most part. Most LLM output can not be copyrighted, because there is not enough human essence in it. Meaning that it is public domain the moment it is produced.

There might be a few exceptions, if you really guided your model in a certain particular way, but that is rare.

What /could/ be is that you violate their ToS. If that ToS is valid and legally and practically enforceable is a different question and will depend a lot on your jurisdiction.

Oh, which book if you don't mind sharing?

That's not a good point though, a lot of countries, including the EU, also subsidize fossil fuels.

Solar is cheaper, but requires more room and time to spin up (think datacenters, where you can put a turbine within a month) and storage or backups for windless nights.

Panels absolutely, their price is more and more dominated by installation costs. You can probably use automation to bring that down as well, especially for large-scale installations. Not sure if that requires AI research though, it's more an engineering and funding challenge.

Batteries are still open. While they do get cheaper, there is still a lot of room to improve. And battery chemistry is something where a lot of research, trial and error, healthy intuition is necessary. I'd say that is more a field where an AI based approach might make sense.

This can be out of your control. Your health insurance will ask why that camel hit you, and might sue the bus operator for their damages for letting you out where they should not.

They don't have access to EUV machines from ASML though.

I never understood why visited websites and all their first and third party cookies are not isolated by default in browsers.

There are so few sites where I really need cross-site logins to work that an opt-in would be much more preferable.

Have all sites set cookies on whichever third party they want, but don't share those cookies over. Add an explicit "Do you really want foo.com to share data with bar.com?" if you really need to see those facebook comments on your news site.

The supply chain both in parts for their machines, but also in terms of research and technology for new machines is very deep and very international. If the US states that a technology can only be used by ASML if they keep R&D in the US, then there is no easy alternative for ASML.

That is true not only for ASML, but for the semiconductor industry in general. The number of smaller, specialized companies for certain tools, chemicals and software is staggering. And they sit all over the US, Europe, Japan, Taiwan and Korea.

The good thing about math is that you can formally verify the results, making hallucinations much less scary.

This is similar to games, where large learning models had some of their first big wins (go). Games, like math, have strict rules that can be used to prune invalid outputs.

It is already in the making, the next big step in math will be the complete formalization of all existing, relevant theorems and proofs via LLMs.

It's one of those "imagine a frictionless, perfectly spherical pig in a vacuum" theories that don't survive contact with reality.

Buyers don't just pop up on timestamp zero and remain unchanged. They anticipate price changes, potential new buyers come in, the market is dynamic.

I also don't understand why this only affects monopolies. The same logic should dictate that all products and services fall towards MC?

Can it in some way do some action based on the talk? For example, can it summarize the brainstorming into a text file, post it somewhere etc.?

Essentially part of a machine vision software package, which in turn is computer vision specialized for factory automation. It's "industrial" because it was trained on such data (more expiration dates and SEMI-font on wafers, less random pictures of streets with some text).

That, plus engineering and support: a robust implementation, deployment on almost any system, fast and good long-term support so your factory doesn't stop for too long if something goes wrong etc.

> OCRing this is a nightmare and is a good benchmark to any self-proclaimed good OCR/vision model.

It's not that difficult, our industrial OCR model read it correctly on its first attempt with default parameters. The characters are easily separable, there is no structured background (think expiration dates on yogurt aluminum lids) that confuses the reader, there is no almost-text-like texture anywhere that would clutter the result. The font is also nice and standard.

Mostly the second point.

They already collect and track you, even with car play. I strongly recommend this CCC talk, where they hacked a Volkswagen database that contained unfiltered, high-accuracy, timestamped locations of a large majority of electric cars from VW group.

Based on that they were, for example, able to identify cars owned by members of Germany's security apparatus: where they work, where they live, where they drop off their children each morning. Who visited brothels.

https://www.youtube.com/watch?v=SGzoXbbth0s&t=30m

It is like questioning the purpose of shovels, because surely if shovels generate so much wealth for the miners, wouldn‘t shovel companies ask for a percentage of the gold (say 30%) instead of selling shovels at a unit price?

The risk is that this does not scale, they run out of money, the robot won't work any more, and you payed a lot for nothing.

Is there a zero-retention option?

No, we get an egg (i.e. the part usually provided by the female), which must then be fertilized by sperm.

I don't know enough to say if that allows extracting eggs from males. Could two males have a child together using this technique?

Bridges deteriorate by default. That is a well understood process with well understood reasons, which we monitor and work against. Salt, thermal cycles, load, rust.

Humanity does not deteriorate by default. Claiming it does so through some hand-wavy pseudo-evolutionary arguments is not a strong case, and requires at least some evidence to be taken serious. How about a (equally unfounded and just for the sake of argument) reverse claim: Humanity got more intelligent, because high child mortality favored physically strong children instead of mentally strong children.

> I had been cured of tuberculosis by them... in 7 days

Just FYI that might suck in the future. Some countries might deny you entry / visa, or at least require expensive additional screening if you had tuberculosis diagnosed in the past. It might be worth to fight this.

The thing is we tried that for decades, using more formal logic to build reasoning engines. And we never got it to be even a fraction as good and generic as learning-based LLMs are today.

This interpretation also generalizes nicely to 3d. You have 60 moves, 20 of which will be up (60 choose 20), then 20 are east (40 choose 20), then 20 are south (20 choose 20).

But were they given access to the raw data, or reports?