At least Oracle Weblogic is being useful for someone.

There’s a part of our brain called the salience network, that continually models and predicts our environment and directs our conscious attention to things it can’t predict. When we talk to each other, most of the formal content is predictable, and the salience network filters it out; the unpredictable part that’s left is the actual meaningful part.

LLMs basically recreate the salience network. They continually model and predict the content of the text stream the same way we do—except instead of modeling someone else’s words so they can extract the unpredictable/meaningful part, they model their own words so they can keep predicting the next ones.

This raises an obvious issue: when our salience networks process the stream of words coming out of such an LLM, it’s all predictable, so our brains tell us there’s no actual message. When AI developers ran into this, they added a feature called “temperature” that basically injects randomness into the generated text—enough to make it unpredictable, but not obvious nonsense—so our salience networks will get fooled into thinking there’s meaningful content.

It’s like eleven Floridas crammed into the space of Michigan.

I have little hope that Biden, Harris, or (obviously) Trump will actually change course on the US’s Israel/Palestine policy—but to be fair, we shouldn’t expect the current vice president to openly say she would reverse the current president’s foreign policies even if she intended to.

“The boat suffered a series of indescribable, unreasonable errors, the impossible happened on that boat … but it went down because it took on water.” […] The CEO ruled out any design or construction errors, which he called unlikely after 16 years of trouble-free navigation.

This sounds like an out-take from the “front fell off” skit.

According to the article, there were restrictions imposed three years ago to pressure the Saudis to reduce civilian casualties in the Yemeni war, and (according the to administration), the Saudis complied.

You know someone messed up when the Saudis feel obliged to defend Iran.

Rather than creating a custom terminal app, could you create a user that only had permission to run the restricted commands, with a profile script that gets run at login and offers a menu of common tasks?

It’s a feature, not a bug: if he claimed to know how it worked, they wouldn’t be able to sell it as a scapegoat for indefensible business decisions.

Does it need to be accessible via API (e.g. SQL) or just a spreadsheet-style web interface?

You can use any port for SSH—or you can use something like Cockpit with a browser-based terminal instead of SSH.

Ok, but protests need to be done “without distraction”? That kind of defeats the purpose, doesn’t it?

The US Navy will convert its surplus oil rigs in the Pacific into active mobile military bases

The Navy has its own oil rigs?

If you didn’t map a local config file into the container, it’s using the default version inside the container at /app/public/conf.yml (and any changes will get overwritten when you rebuild the container). If you want to make changes to the configuration for the widget, you’ll want to use the -v option with a local config file so the changes you make will persist.

So they’re using xenophobia to attack education, because xenophobia for its own sake isn’t evil enough.

“Wikipedia” —Lol
“US officials claim” —Lol

…Says the user who just posted a story sourced from Russian state media.

(Not that the report is necessarily wrong, or that US claims are necessarily true—but you’re applying a bit of a double standard with your instant dismissal.)

“Feedback training teaches models to consider their racism,” says Valentin Hofmann, a researcher at the Allen Institute for AI and a coauthor on the paper. “But dialect prejudice opens a deeper level.”

Hmm… I think dialect bias is a distinct problem, which may need a separate approach that doesn’t just lump it together with racism and try to eliminate both using the same means.

I just quickly skimmed the source paper, but I’m wondering if they can distinguish between a latent “concept” space that’s actually based on English, versus one that’s derived from all source languages and just weighted more towards English?

“I like it when companies game search engines by titling their promotional pages to mimic search query results.”

The Mac keyboard layout was largely inherited from the Apple ][ and pre-OSX systems, which were non-Unix. The original Unix keyboard layout differs from both Mac and Windows layouts.