954 karma · joined April 20, 2023
Not really.
>I wrote that SQL schemas would provide “only marginal value” to an attacker. Big mistake. Chicago jumped on those words and said “see, you yourself agree that a schema is of some value to an attacker.”
The City of Chicago's argument was that something of ANY value, no matter how insignificant, would help an attacker exploit their system, and was therefore possible to keep secret under the FOIA law.
What fresh hell is this?
That's a terrible use for an LLM. There are several deterministic search engines attorneys use to find relevant case law, where you don't have to check to see if the cases actually exist after it produces results. Plus, the actual text of the case is usually very important, and isn't available if you're using an LLM.
Which isn't to say they're not useful for attorneys. I've had success getting them to do some secretarial and administrative things. But for the core of what attorneys do, they're not great.
This is also my experience. My day job isn't programming, but when I can feed an LLM secretarial work, or simple coding prompts to automate some work, it does great and saves me time.
Most of my day is spent getting into the details on things for which there's no real precedent. Or if there is, it hasn't been widely published on. LLMs are frustrating useless for these problems.
For some tasks they're still next to useless, and people who do those tasks understandably don't get the hype.
Tell a lab biologist or chemist to use an LLM to help them with their work and they'll get very little useful out of it.
Ask an attorney to use it and it's going to miss things that are blindingly obvious to the attorney.
Ask a professional researcher to use it and it won't come up with good sources.
For me, I've had a lot of those really frustrating experiences where I'm having difficulty on a topic and it gives me utter incorrect junk because there just isn't a lot already published about that data.
I've fed it tricky programming tasks and gotten back code that doesn't work, and that I can't debug because I have no idea what it's trying to do, or I'm not familiar with the libraries it used.
If there's a lot of tryptophan in the bacteria, the ribosome scoots along the mRNA quickly and a kink forms later in the mRNA that boots the RNA polymerase off of the DNA prematurely, and stops the cell from making the mRNA that encodes for proteins that make tryptophan.
If there is a deficiency of tryptophan in the cell, the ribosome stalls, and a different kink forms in the mRNA that allows the RNA polymerase to keep on transcribing the genes necessary to make tryptophan.
This only works in bacteria, because there's no nucleus, and transcription and translation can occur simultaneously.
Offering someone credit in USD isn't making your own currency. The closest widespread version of that are community currencies[0], which the US doesn't seem to particularly care about -- I'm guessing because they're generally pegged to the dollar and promote local economies.
[0] https://en.wikipedia.org/wiki/List_of_community_currencies_i...
Price instability, confiscatory and variable transaction fees, several high profile frauds -- including in a so-called "stable coin".
Crypto is its own worst enemy. Not the government.
That appears to be the premise of DOGE.
This sentence also killed me:
>working on various projects I’m definitely not able to talk about
It's not because they're classified. It's because a bunch of unaccountable private citizens are plotting behind closed doors to tear down the government.
This is more an anti-trust problem than deregulation. In the 80s Robert Bork and some others led a charge within the court system to rewrite federal anti-trust law without actually rewriting the laws. The result was that merger/acquisition guidelines were loosened, and the focus was on whether the result would cause the consumer to pay more money. And thanks to the way health care is paid for, it's a little complicated to make that case.
Intel is currently investing $7B a quarter into getting their foundries competitive again, and it's not clear yet that they'll be able to really do so at scale. And even if they do, it's not clear whether those foundries can effectively serve customers that aren't Intel.
The reason people trust TSMC to make their chips is because TSMC isn't making a competing chip. If I come to an Intel Foundry with my design and work with them to spin up some new capability to get the features to work right, there isn't much of anything that stops Intel Chips from using that new capability to compete with me in a year.
https://www.pcmag.com/news/intel-developing-ai-chips-for-ama...
Intel is being weighed down pretty dramatically by trying to do 5 nodes in 4 years AND do it at any scale. Intel's foundry business lost 2.8B last quarter, on revenue of 4.3B. That means they spent 7.1B in THREE months.
Intel currently can't use any of its new nodes to make their leading chips (AI PC, e.g.) because those nodes aren't operating at scale. So they're paying TSMC to make the chips. Which means that Intel isn't sending work to its own fabs to offset the costs of advancing nodes.
Plus, the foundry business hasn't managed to announce any flagship deals with outside parties.
By making it a sub at least they can raise funds to burn someone else's money while they try to spin things up.
The bigger headwind they face is that if Intel is tightly tied to the fabs, those fabs are going to be focused on serving intel, and not on serving the other customers it needs to service to have a leading node fab be remotely sustainable.
Bugs in beta software?
Seems like it's an alternative to tk/tkinter.
CUT TO:
"I'm not a fan of strlcpy(3)"
I saw a video of someone making a ring out of polycrystalline diamond and…if that gets stuck for some reason, it's never coming off.