Until you work in financial markets and you want to measure latency vs a bunch of other servers you don't own, all of whom pick different smears (and some of whom aren't sure in advance which smears).
Assuming your labour contribution to these agents is 'minimal':
Why would you own them, instead of some well capitalized billionaire?
To the extent that you do have capital, why do you assume that your 'minimal direction and guidance' would outcompete a full time specialist working for that billionaire?
I'm pretty skeptical of the market argument - reading the linked article, it seems low granularity subsidizes market makers when there are many small orders (which often get worse prices than they would otherwise) and this subsidy allows market makers to be deeper, which subsidizes large orders. Which is a difference, but not an unalloyed good.
Surely price competition is more economically valuable than the queue position game, which is measured in economically meaningless nanoseconds.
If you run an interview process where candidates who take 6-8 hours and claim to have taken 4 hours score highest, those are the candidates you will hire.
This is capitalism working as intended. Only the best run airlines can survive, and investors are collectively subsidizing air travel for non-investors.
I think AST aware code reading is criminally underused by agents - you don't need a header file if you can see a listing of all the functions in a library.
Similarly, I don't read the whole file a function is in while editing it in an IDE, why should a coding agent get the whole file polluting its context by default?
This is interesting, but doesn't have to be correct.
If Blockbuster had kept pouring money into the new service, maybe it would have lost it all - I see no reason to think Blockbuster's movie rental franchise business would have 'transferrable skills' to allow it to succeed at streaming.
If it had been trying to pivot into a pizza delivery business (perhaps more transferable, in terms of locating franchises etc) would Icahn still have been 'killing' it?
My point is, maybe it was already dead and Icahn just prevented it from wasting a lot of money on the way down the drain.
If it's lasted 10 years and someone is still using it after all that time, that seems like a pretty good signal there's a lot of value in the 'garbage'?
I've seen a lot of 'fixes' for 10 year old 'garbage' that turned out to be regressions for important use cases that the author of the 'fix' wasn't aware of.
If markets were regulated to trade in coordinated 1s auctions, instead of nanosecond precision first-come-first-served matching of orders, markets would function just as well without needing a ton of what the HFT crowd does. It's a massive waste of brilliant minds.
That seems like a legal question - if the model weights contain an encoded copy of the copyrighted material, is that a 'copy' for the purpose of copyright law?
If there’s anything I would want to run to verify, I ask the author to add a unit test. Generally, the existing CI test + new tests in the PR having run successfully is enough. I might pull and run it if I am not sure whether a particular edge case is handled.
Reviewers wanting to pull and run many PRs makes me think your automated tests need improvement.