Three side remarks about China, which can be a peculiar example to compare to for Russia, maybe even any other country:

• They actually banned consoles for a quite significant 15 years (2000–2015), which strongly tilted their market towards PC.
• Their companies actively make PC-type gaming handhelds, and many of them are even well-established in the business ahead the current “Steam Deck” wave/bandwagon: GPD (once called GamePad Digital, first release in 2016), OneXPlayer (2020), Ayaneo (2021).
• Chinese gaming companies are quite at the whim of the censorship, and occasional “crackdowns” out of the blue, and many have therefore reoriented themselves for an international audience to de-risk their business.

Just for reference, a few years back, (ex-Microsoft) David Plummer had this historical dive into the (MIPS) origin of the blue color, and how Windows is not blue anymore: https://youtu.be/KgqJJECQQH0?t=780

Likely due to being a prototype. Production laptops from Tuxedo tend to have the “TUX” penguin in a circle logo on the Super key by default. They also have been offering custom engraved keyboard (even with the entire keyboard engraved from scratch to the customer’s specifications) as added service, so probably there will be suppliers or production facility to change the Super key.

By the way, there was one YouTube channel that ended up ordering a laptop with Windings engraving from them: https://youtu.be/nidnvlt6lzw?t=186

How does this analogy work at all? LoRA is chosen by the modifier to be low ranked to accommodate some desktop/workstation memory constraint, not because the other weights are “very hard” to modify if you happens to have the necessary compute and I/O. The development in LoRA is also largely directed by storage reduction (hence not too many layers modified) and preservation of the generalizability (since training generalizable models is hard). The Kronecker product versions, in particular, has been first developed in the context of federated learning, and not for desktop/workstation fine-tuning (also LoRA is fully capable of modifying all weights, it is rather a technique to do it in a correlated fashion to reduce the size of the gradient update). And much development of LoRA happened in the context of otherwise fully open datasets (e.g. LAION), that are just not manageable in desktop/workstation settings.

This narrow perspective of “source” is taking away the actual usefulness of compute/training here. Datasets from e.g. LAION to Common Crawl have been available for some time, along with training code (sometimes independently reproduced) for the Imagen diffusion model or GPT. It is only when e.g. GPT-J came along that somebody invested into the compute (including how to scale it to their specific cluster) that the result became useful.

This is a very shallow analogy. Fine-tuning is rather the standard technical approach to reduce compute, even if you have access to the code and all training data. Hence there has always been a rich and established ecosystem for fine-tuning, regardless of “source.” Patching closed-source binaries is not the standard approach, since compilation is far less computational intensive than today’s large scale training.

Java byte codes are a far fetched example. JVM does assume a specific architecture that is particular to the CPU-dominant world when it was developed, and Java byte codes cannot be trivially executed (efficiently) on a GPU or FPGA, for instance.

And by the way, the issue of weight portability is far more relevant than the forced comparison to (simple) code can accomplish. Usually today’s large scale training code is very unique to a particular cluster (or TPU, WSE), as opposed to the resulting weight. Even if you got hold of somebody’s training code, you often have to reinvent the wheel to scale it to your own particular compute hardware, interconnect, I/O pipeline, etc… This is not commodity open source on your home PC or workstation.

The situation is somewhat different and nuanced. With weights there are tools for fine-tuning, LoRA/LoHa, PEFT, etc., which presents a different situation as with binaries for programs. You can see that despite e.g. LLaMA being “compiled”, others can significantly use it to make models that surpass the previous iteration (see e.g. recently WizardLM 2 in relation to LLaMA 2). Weights are also to a much larger degree architecturally independent than binaries (you can usually cross train/inference on GPU, Google TPU, Cerebras WSE, etc. with the same weights).