It's free for now, just like registries were "free" and docker desktop was free.. until they weren't. I am not against Docker capitalizing and charging for their services (as they should); however, the pattern of offering a service for free and then reneging after it's widely adopted, makes me hesitant to adopt any of their offerings.
A month or two after a visit to Iceland (my favorite country by far by the way), I received a ticket in the mail for speeding. It included a picture of the car I rented and a closeup of the driver's face- a face that did not belong to me (presumably another renter).
Luckily, a quick phone call and a copy of my drivers license cleared things up, but systems like these inevitably lead to "guilty until proven innocent" scenarios instead of "innocent until proven guilty".
From the abstract, it sure sounds like any electronic checkers or chess game would fall under this patent. If so, I'm sure there is plenty of prior art to invalidate their claim.
I'm not even sure it saves a ton of time to be honest. It sure _feels_ like I spend more time writing up tasks and researching/debugging hallucinations than just doing the thing myself.
Everson on the @GeekDetour YouTube channel explores how a patent filed in 2020 could be preventing our slicers from generating stronger 3d prints using "Brick Layers". He shows, what appears to me at least, to be pretty clear evidence of prior art that _should_ invalidate the patent.
I'm not a lawyer but would love to see this added to all slicers.
It sounds like the same requirements as a 70b+ model, but if someone manages to get inference running locally on a single rtx4090 (AMD 7950x3D w/ 64GB ddr5) reasonably well, please let me know.
This is not always possible and a little short-sighted, especially when you are on a large, complex, established project with many contributors, where "just refactor 1.5 million lines of code, spanning 1500 files that have evolved over a 10 year period" is just something not open for debate.
In those cases, and in the cases where a single large file just makes more sense or is unavoidable (IaC/build scripts, shell/sql/migration scripts, config files, etc.), the author's methods are absolutely valuable.
I find if you subscribe to the "Write libraries not software" mantra where your libraries/packages/classes are focused and adhere to the single responsibility principal, code organization kinda sorts itself out. But I agree with the author that organizing and commenting long files like this makes it easier to grok what's going on.
But if you are clicking around a project and scrolling with your mouse, instead of taking advantage of your IDE's ability to use keystrokes to find and open files, jump to references/definitions, find and replace text, etc., you are handicapping yourself. It's worth the investment to learn.
I really like the idea of something like vlang (https://vlang.io), as I've tried multiple times to find the motivation to learn Rust, but always end up going back to golang for the simplicity.
I realize how much effort is going into a project like this and I do hope you find a way to monetize on it, but until it's open sourced and auditable, I personally could never see me using it. The risk is simply too high.