This is the right approach, but there’s lots of complexity around transcoding. In the courtroom that’s less of an issue if you can get the original unmodified outputs, but broader applications need to think through what it means to be “verified”
As always, the devil is in the details. They defined an AI system as software that uses any of:
(a) Machine learning approaches, including supervised, unsupervised and reinforcement learning, using a wide variety of methods including deep learning;
(b) Logic- and knowledge-based approaches, including knowledge representation, inductive (logic) programming, knowledge bases, inference and deductive engines, (symbolic) reasoning and expert systems;
(c) Statistical approaches, Bayesian estimation, search and optimization methods.
That could very reasonably be interpreted to mean all software. The act is mostly concerned with high risk applications. These are based on intended use, which is obviously tricky in the context of general purpose LLM’s. Most of the covered intended uses make sense, but in combination with their definition above it could mean that, for example, any developer of EdTech software in the EU could be regulated under the act.
More broadly, the act was clearly not written with general purpose foundation models in mind. For example, Article 10 requires that all training data be “relevant, representative, free of errors and complete.” That is pretty much impossible to comply with if you’re training on The Pile or similar.
As someone with neither the background nor the mathematics to fully understand renormalization, I'll say that the wiki entry isn't terribly helpful:
"Renormalization, in this example, mathematically replaces the initially postulated mass and charge of an electron with the experimentally observed mass and charge."
That sure sounds like adding an arbitrary constant into your theory and turning the knob until the theory matches the results. The comments here suggest that's not the case, but it's easy for me to see where the skepticism comes from.
Artifacts have been periodically salvaged from the wreck between 1987-2000. RMS Titanic Corp now has title to those artifacts and has periodically talked about returning to the wreck to salvage more. The whole thing has been tied up in court cases since the 1987 expedition, which is par for the course with international maritime salvage.
The most interesting argument that I hadn't previously seen is that Musk's alleged breach of the agreement (non-disparagement, witholding consent, etc) may prevent him from terminating the deal even if Twitter has violated it (which they of course deny). Of course you have to take that with a big grain of salt given the source of the argument, but it's new to me at least.