The word sardonic used to mean what we now use sarcastic for — verbally ironic. Sarcasm comes from the Greek “to tear flesh, bite the lip in rage, sneer” and meant “bitterly cutting or caustic” when it first entered English. For me, although I understand that hypothetically you could have sarcasm that doesn’t have this inherently negative bent to it, the word still retains a fair bit of its original connotation for good reason.
I’m not as quick as you. I got most of the way through article and was still wondering why X would expose a database of historical prompts to an llm for querying by law enforcement.
I mean, most llm makers work pretty hard to conceal the system prompt, and I have no idea why XAi would give Grok access to a database of historical prompts. LLMs don’t have memories by default, and their inability to learn from past experiences is kind of a big stumbling point for a lot of folks. You can ask, but I doubt you’re likely to get anything other than a confabulation.
I think I like the draft headline better, despite it’s clunkiness.
People commenting after only reading the headline and not the article is exactly the behavior I find irritating and distasteful about headline-related complaints.
I’m usually against complaints about poor headlines, but this one is completely factually incorrect? The FBI didn’t interact with Grok here literally at all? They issued a search warrant to X to get their logs?
deleted by creator
Usually I roll my eyes pretty heavily at the WhateverPorn names for non-pornographic communities, but I suspect this one’s gonna be accurate more often than not.
If you’re comparing freestyling (assuming it’s fully improvised) is this still the case? What about reciting a memorized poem? (I agree with you, I’m just curious if you know where the boundaries are because I sure don’t)
LessWrong is a deeply strange place and you have to be careful with it, but the sequence on maps and territories is a pretty good introduction: https://www.lesswrong.com/w/map-and-territory
My condolences for your loss.
Capricious is my favorite adjective, rather than trying to ascribe mental disorders to global systems. If you’re asking regarding a specific human, you should talk to a psychiatrist — if it is disruptive to their life, it may be a sign of borderline personality disorder.
This use of bipolar has very little to do with the experience of most bipolar people. Bipolar rapid cycling is defined as four episodes per year, much closer to the change of the seasons than the change of weather from day to day.
Not so. The code may be the same, but the code is the map, not the territory. The code is not enough to understand the territory, even with mock data, c.f.
Our staging environment contains data that matches Production as closely as possible, but was not sufficient in this case and the mock data we relied on to simulate what would occur was insufficient.
Ugh. Thanks for the heads’ up — I’ve definitely posted archive links without noticing they’re blocked before. PBS and NPR have really gone downhill with the budget cuts. ProPublica is great, but their coverage is pretty narrow, so there’s a lot of stories they don’t cover at all. It’s getting harder and harder to find a quality source.
I’m certain they’ve wanted to do this for a long time, and AI is a convenient way to justify it, rather than admitting they don’t want humans using it to circumvent the paywall. It does solidify for me personally that the LA Times is the paper of record for the United States going forward, rather than the New York Times.
Any good archiver will check for an archived copy before making a request, and batch requests. This was very different than the attack you’re imagining — if you opened any archive.today page, it would poll a developer’s personal blog, regardless of whether you were interacting with content from that blog.
Unfortunately, they’ve allegedly modified the contents of some archived articles, so even though they may do better to archive, nothing archived is of any value because it cannot be trusted.
It does more to handle client-side rendering than archive.org, so there are pages that could be rendered by today that were not archivable by org. Also, because of differing usage patterns, it has archives of pages that org didn’t, and even for pages that org does have, at times org doesn’t.
I think unfortunately we’ve lost a lot of that older meaning as it’s been crowded out by a more racially charged usage.