stvltvs

I'm responding to the push to rebrand it as Super Intelligence coming from people in power (not just the current occupant of the White House). Whether or not they believe it to be super human, they seem to want us to see it that way.

How would we train those LLMs? Feed them binary executables along with their business requirement docs? Is the corpus of training data large enough to be effective? What happens when a new CPU architecture is developed? Where would the training data come from for the new machine code?

Exactly, at what point do spec files become so precise that they are just another form of source code?

It's almost like LLMs are another tool with strengths and weaknesses rather than a super human intelligence.

As anyone in the Rust Belt can tell you, the average worker might experience higher wages over time, but the average worker is a statistical fiction, not a flesh and blood person.

Individuals and entire communities can live out the rest of their lives destitute and hopeless in the wake of changes that leave them behind. Things might be brighter for those who come after, but let's not pretend we're not doing real violence to those left behind.

What other industries have enough vacancies to be filled by refugee software engineers without significant retraining or cuts in pay?

My local utility has basically this program.

What are the chances that it's deployed at the Canadian border?

Focusing on the specific death rate doesn't change that crossing the border is risky, and some people will take their chances with or without the heat ray. It's one more variable in the calculation.

Wikipedia is available if specific mortality rates matter. https://en.wikipedia.org/wiki/Migrant_deaths_along_the_Mexic...

Every individual has their own risk tradeoff calculation. People already risk death to cross the border, so it's not clear how much this would move the needle.

Or reckless.

I'd say that "eat less" and "make a lot of money" are both helpful, winning strategies but lack the tactical details necessary to bring them into effect. Agreed that it's not helpful to toss them out at someone as if implementation is easy.

There's also a huge asymmetry between what clients know about each other and what the platforms know about the clients. Bot detection is only really feasible by the platforms.

I'm old enough to remember the days before online porn, and all I can say is that sometimes a difference in quantity is so large that it becomes a difference in kind.

+split screen videos

What portion of grand conspiracies do we think go undetected forever? Do we have any reasons for our intuitions?

It's free, and I'm boycotting Amazon.

In cases where I have trouble getting a book from the library, it's often on Spotify. An Audible subscription would be mostly redundant.

Agreed that they're overpriced. For some reason, I rarely want to own a DRM-ridden audiobook, especially at those prices. (The Lord of the Rings as read by Andy Serkis and The Buffalo Hunter Hunter are two recent exceptions that you should check out.)

Once I'm done borrowing an audiobook from the library through Libby or Hoopla, I might buy a physical book if I liked it enough. I usually don't foresee myself listening to the book again.

Big Labor j/k

There's a deeply ingrained mythos around how natural and easy motherhood is, but in nature, a lot of first time mothers are rather mediocre. It takes a while learn how to do all those "natural" things like nursing the baby.

And let's not forget that babies kinda suck (intended) at doing "natural" things too.

I can't vouchers for the specifics, but it could make sense because gas and electricity prices as well as incentives vary a lot between US states.

The video is more widely accessible on YouTube https://youtube.com/watch?v=KvanHITPNEc

And we don't realize that we are the poor just because we can point to poorer people.

FWIW I don't agree with Wilhoit, just explaining where the idea came from.

Which is basically Wilhoit's Law, an attempt to define conservatism.

edit: For those who don't want to look it up, Wilhoit's law is usually stated like this: "Conservatism consists of exactly one proposition, to wit: There must be in-groups whom the law protects but does not bind, alongside out-groups whom the law binds but does not protect."

Because they know that many of our most intractable problems are primarily political, not technical. Climate disruption and peace in the Middle East for two examples. I don't expect an LLM to offer a workable plan for either. Would love to be wrong about that though.

Trump et al. being in control of US government data honestly feels like vikings have been put in charge of the Anglo-Saxon monasteries or Christian zealots in charge of the Library of Alexandria.

That's not the story the whistleblower tells. The plan, for example, is to reject entire batches of ballots if a single ballot doesn't pass USPS's criteria.

The idea is to be able to prove that a photo you took was captured by a camera, not generated. That would be helpful in the context of a disputed news story, a court case, etc.

Would that ever be relevant for a screenshot?

Signing the raw image data wouldn't require also signing the EXIF metadata. For privacy, you could later strip out everything except the raw image and the camera's signature and still prove that the image is unaltered.