zogwarg

Also used for practicing my Japanese, I would say the usefulness was definitely not 0, if nothing else with the correct settings it was daily practice for the Japanese scripts (kana/kanji).

And it's best It would definitely not bring you even to reading fluency, and was only good if you were supplementing your study with other language acquisition forms (like for example, in my case, living in Japan).

The examples were often stilted, and the accepted answers overly rigid, for sentences which weren't necessarily realistic.

I think some of the worst aspects of the gamification were:

• Choosing easy exercises as a safe source of points, to not lose the streak. (perverse incentive)
• Essentially by setting a target, encouraging to only meet a daily points streak, and not necessarily go further for a given day. (perverse incentive)
• Tile matching to english, again with overly rigid accepted answer. (trying unhelpfully hard to map Japanese to english)

Let my streak/subscription lapse when it stopped being useful (got better reading exercise elsewhere), and uninstalled when they introduced AI shit.

Also I think there's enough manipulation fantasy in HPMOR, and enough lack of agency from Hermione, that it qualifies—in it's own way—implicitly as being erotic.

Also I realize the word often get's used fuzzilly that way even in general, but I suspect what they mean is epistemology not ontology.

You have to wonder about that Tim traveler; Merlin?

For reference the open-source project appears to be OpenCog, founded by "Ben Goertzel" who at least up until 2010 held the title of "Director of Research" at SIAI, the relation stopping because he wasn't a true believer in doom.

Honestly even the original paper is a bit silly, are all game theory mathematics papers this needlessly farfetched?

I can't imagine anyone really subjecting themselves to reading all that, I'm delighted for them though, or distraught that it happened...

It is a bit sad how Yarvin frontlines his "victory", by quoting some extruded text, but in context—he is somehow kind enough to provide, maybe he didn't bother reading all of that either—it's just some fence-sitting big nothing, i doubt the claims that this produces any form of "red-pilled" Claude.

(I'm not sure what I expected, but it truly was a dead dove.)

• AI Naughty List
• AI Slop Watch
• AAI = Avoid AI Index = Anti AI = Another AntiAI Index

Difficult difficult, too jet-lagged.

🫂

I think actually listening to people remains important. But you're only truly listening to someone when you try to understand when they lie or the ways they can be wrong.

Assuming 100% good faith is not actually the most empathetic way to engage with a person.

A get-smart-quick scheme, with an absurd amount of occult cargo-culting of scientific discourse.

~~In the dead of night, knocking on the door~~