HN user

thebooktocome

1,613 karma
Posts0
Comments826
View on HN
No posts found.

To take a simplistic example, even janitors might potentially have access to ITAR controlled materials if engineers threw them out into a bin bound for a shredder/incinerator and janitorial staff are trained and trusted parts of the disposal chain (which they should be!). Anyone might be able to overhear water cooler conversations.

As I mention elsewhere in this thread, if this is really how SpaceX operates then they’re already in (ethical, if not legal) violation of their duty to protect ITAR from dissemination. ITAR documents should never just be sitting in an unsealed bin waiting for disposal. Employees shouldn’t casually be discussing ITAR around the water cooler.

Assuming everyone around you is permitted to handle ITAR is a recipe for disaster.

For SpaceX to internally segment ITAR from non-ITAR is a huge bureaucratic overhead for them which leads to a possibility of mistakes.

They decided to become an aerospace engineering firm in the US. ITAR security is part of the cost of doing business.

If the typical e.g. janitor or cafeteria worker at SpaceX has access to ITAR, as SpaceX seem to have alleged before they got caught, then their ITAR security is pure theater.

SpaceX recruits and hires for a variety of positions, including welders, cooks, crane operators, baristas and dishwashers, as well as information technology specialists, software engineers, business analysts, rocket engineers and marketing professionals. The jobs at issue in the lawsuit are not limited to those that require advanced degrees.

The fact that SpaceX deals in ITAR does not prevent them from hiring refugees for roles that do not handle ITAR.

The client "needs" in question here are low cost overall, low marginal cost for each revision, and a totally-interactive "do what I mean" interface.

Shoving a human artist in the middle is a liability on each front.

Can you give an example of a “clear and bright” line in copyright law that does protect “folks who don’t have the money to fight an uncertain battle in court”?

For context, I’m in the process of translating a work that I know for a fact is in the public domain (sole author died 90+ years ago) and I’ve still got legal questions that I’m going to have to hire a lawyer to solve.

Overall I think this is a net win, especially because I don't think this needs to be the end of the road for human illustrators, but this will force them to adapt and bring more sensitivity to the needs of their clients.

The advantages of AI that you crow over simply can’t be met by any human professional artist. A human can’t do hundreds of revisions profitably. There’s increased “sensitivity” and then there’s needing to read the client’s mind.

If you think this isn’t a death knell for human illustrators in this particular market, you’re deluding yourself.

Author conflates legal and ethical options for preventing copyrighted work from being used to train ML image generators.

There’s nothing legally inconsistent about passing a law saying, e.g., “ML training is not fair use”. Doing so will not even reduce existing fair use rights being exercised by actual people.

The author’s argument is that doing so is philosophically analogous to human creative processes, but those are—and I can’t underline this enough—human. And the law is not (and cannot be, should not be?) consistent in such a way.

I don’t think she’s substantially lip-syncing (maybe here and there, but not continuously). I was at the Detroit show and she confessed to having a bit of a cold during the interim into Evermore. And indeed, she coughed and sniffled a bit during other songs it was picked up by her mic. She also got a little short of breath during heavy dance numbers.

Maybe I’m easily fooled, but if she was lip-syncing a lot I don’t think they’d leave these imperfections.

I’m saying that on the continuum of Marxism, the PRC circa 1965 is somewhere to the middle-far-left and the article above (that OP claimed was Marxist) is all the way on the far right.

I’m not super interested in whether or not their Marxism was “authentic”, and neither is anyone who goes around, as OP did, randomly labeling things they don’t like as Marxist.

It’s the same folk-interpretation of Marxism that always goes around. Now that Marxism is dead and can’t defend itself, it’s whatever anyone wants it to be.

States that were far closer to Marxism, like the PRC during the Cultural Revolution, decimated their universities and dispersed their faculties over the countryside, or to labor camps.

This article is not on the same planet as that madness.

It’s not related to the article, but the UK is currently restructuring its farming subsidies to, among other things, encourage sustainable farming and mitigate food security risk.

Basically everyone is unhappy about it for some reason. Some sheep farmers were told to reduce their flock by 90% or so.

But on the other hand, the pastures they’ve traditionally used to graze can no longer sustain such flocks (although this is controversial, it seems borne out by my reading). So if they continue to consider only profit motives, the pasture will end up dead and they’ll be out of work anyway.

It’s a hard problem that ought to have been solved decades ago.

I’m not the parent but I imagine they’re referring to e.g., some FFTs use different partitioning strategies in different threading environments, which breaks bit-perfect replication.

There’s also the weirdness that in C++ the floating point environment is thread-local, which can cause all sorts of chaos.

Where did I imply that? The fault was with the JVM.

“The impetus for changing the default floating-point semantics of the platform in the late 1990's stemmed from a bad interaction between the original Java language and JVM semantics and some unfortunate peculiarities of the x87 floating-point co-processor instruction set of the popular x86 architecture.”

Strictfp is no longer used as of JVM 17. IEEE-754 semantics are default.

https://openjdk.org/jeps/306

Ironically, using your system’s math libraries will probably make replicating floating point results harder. Especially on macOS. Accelerate is a curse.

I disagree with your gloss of Kahan’s philosophy. His approach is more along the lines of “do not waste precision”. But this philosophy is not the complete truth; as close as I can state it briefly, my modification would be “do not waste precision that may be needed later”.