Giefo6ah

Even in the 19th century you didn't need to be a prophet to ponder what will happen when machinery finally becomes fully automated, and what the process of arriving to this state might look like.

Indeed this is the main question of Marx's work.

1984 has been out of copyright in most of the world for 6 years and that, if anything, has made it easier to find an unmodified copy.

If you live in the USA, the "decoy notes" may be considered an "effective technical measure" from the "Digital Millennium Copyright Act".

If you live in Europe, this restriction may be considered "gatekeeping" and exempted by the Digital Markets Act.

Don't bring attention to yourself by asking for permission. Publish your codec, and if the company cares about this they will send a cease and desist.

If you want the world to benefit from your code but you don't want to be responsible for it, try to adapt the codec to ffmpeg. The ffmpeg project is used to dealing with these matters, and will keep your codec working for eternity.

> Why would speakers be any different than earphones

There are various "ideal" EQ targets for earphones depending on what your model of an ideal audio source is. They all start with a flat response speaker, from then you can model:

- Two speakers right next to the listener's ears. (flat)

- A single speaker 1 metre away in the x axis, in an infinitely-sized anechoic room. (free field)

- Two speakers 30 cm away pointing to a plastic torso at a 90° angle, in a room with reverberations. (B&K/Harman)

- A spherical room lined with speakers, point listener in the centre with no reverberation. (diffuse field)

- An infinite wall of speakers 1 metre away in the x axis. (another method for free field)

- etc

Subway (the fast food chain) calls them "30cm subs", or colloquially "a thirty".

The smaller sandwich is a "15cm sub", or "a fifteen".

I'd bet this is what she meant because I've done professional translation in the past.

The input and output of a GPT it not a sequence of messages in turns: fundamentally the input is the entire context (both from the user and the model), and the output is a probability distribution for 1 single next token.

The chat is only one of many interaction patterns possible with that primitive: one where the "most likely" output token is concatenated to the previous input and this is repeated until a limit is reached or the model produces a special "end of sequence" token.

Representing a more complete interaction pattern as a chat is impossible because the chat abstraction is lossy and could never represent it. Sometimes the abstraction leaks, as with the concept of "temperature", that's not a parameter of the model but of the chat.

Deepl's professional advantage is not in the translation itself (even though it has always been a notch above the rest), but in the UI. You can literally click any word on the translated text and (a) see a list of alternatives, and (b) pick one and then the following text adapts to your choice.

In technical terms it lets you stop the transformer at any token, sample alternatives from the probability distribution at that point, pick your desired next token and resume the generation.

This UI lets an expert guide and control the output of the tool, instead of having to fight it. I wish there was a frontend for large GPTs that worked like this instead of the ubiquitous-yet-useless chatbot.

You're mixing a house as personal property with the much broader claim of private property on land.

It is possible to protect those two uses you mentioned, and other reasonable uses of land, without at the same time spending state resources in enforcing perpetual, absolute, hereditary, exclusive titles on plots of land.

The cones are not sensitive to a single wavelength but to a range.

The green-sensitive cones overlap with the red-sensitive cones, and to a smaller extent also with the blue-sensitive.

Full saturation red and blue are possible by emitting light on the edges of the visible spectrum.

Full saturation green, however, also activates the red and blue cones.

To cover the whole gamut is impossible, but you can approximate it with ~three green tones: a 490nm deep cyan that hits blue and green but not red, ~510nm that hits red and blue equally, and ~540nm the peak of the green cone.

When the model is open weights you can even pass every token (including the chain of thought) though a fourth-party lightweight model like gpt-oss-safeguard to check that it has not become adversarial.

They clearly didn't have the compute capacity ready to release Fable at any price that would have seemed "worth it".