Jordan117 @ Jordan117 @lemmy.world

Posts

133

Comments

304

Joined

2 yr. ago

Moderating

---

This reminds me of when the Ukraine blackmail story first broke. Even before the impeachment push, it seemed like the story was far more riveting -- entirely because of the whistleblower angle. I remember thinking at the time that if all the transcripts being shown on the news with the reactions and dramatic highlighter effects had instead been a public statement by Trump that "yeah, we're withholding weapons until Zelensky gives us dirt on Biden," the entire thing would have blown over in a week. Maybe this could be similar?

(You know what else has the cachet of secrecy and coverup? The Epstein files.)

So sick of these kinds of posts. At this point it should be obvious that the blatant hypocrisy isn't stupid or some clever "gotcha," it's the entire point. It's a demonstration of impunity and contempt for fairness and logical consistency.

Relative to the speed shown in the visualizations. The pop-sci animations look like it takes one step every second or so, while in reality they do hundreds per second.

One of these days AI skeptics will grasp that spelling-based mistakes are an artifact of text tokenization, not some wild stupidity in the model. But today is not that day.

How is it more accurate? I thought that in reality the "steps" are incredibly fast, but that makes it seem like it's a much slower and more uncertain process.

It's Twitter, so there's an excellent chance this is manipulative engagement bait. But it would be really, really interesting if heavy usage of AI image gen proved to be an effective kind of visualization "exposure therapy" for people with aphantasia. We've never before had the ability to so quickly and reliably convert words to images, so maybe experiencing that connection on demand a few thousand times is enough to activate those mental pathways for people who lack them? The closest we've had up to now is a Google Image search, and those results are much more varied, not as precisely tailored to the search term, and not something that people generally do over and over again for leisure.

I suspect a large chunk of people in denial have either never used it and are basing their judgment on memes and cherry-picked failure modes that go viral, or have only used the earlier/weaker models that are far more prone to hallucination. I use ChatGPT Plus regularly, and find it occasionally imperfect but still very useful, but a few times I've tried using it while logged out and the free model it defaults to was far more flawed. Probably why OpenAI built GPT5 around a marginally better but far more efficient model that they can roll out to everyone (including the large cohort of free users).

You can make the argument that AI is damaging or unsustainable or not worth the cost, but acting like no one likes or uses it is straight-up delusional. (Not as delusional as people falling in love with their chatbots, but the fact that people are kind of proves my point.)

Not everywhere! This fucks over anybody who depends on ratings for job evals.

I'm wary of the hype over this gambit because it's been tried several times in the recent past and has always ended with a few key legislators caving and returning to the capitol for no real concessions. Hopefully they realize that people are not in the mood for token resistance (and Abbott's threats against them will steel their spines).

I have -- I had high hopes for it, but the vocals are still pretty bad. They all have this metallic, tinny quality, liked they're subtly vocoded or being sung into those little toy spring microphones for kids. It's a constant reminder that it's artificial and just completely takes me out if it.

I don't know how Suno has become so much more popular than Udio. Every Suno track I've heard has sounded like the same generic pop, and the vocals always have this noticeable "synthy" quality.

What are you talking about? Yang dropped out before the CA primary, and Harris dropped out the December before the first primary happened. She wasn't even on the ballot.

GEORGE IS IN WEIRD TROUBLE!

At first I thought the quoted text was just him having ChatGPT write his tweets for him. But then I clicked through and saw the video of him actually speaking it to the camera. He was probably just reading it off a screen, but seeing a dead-eyed person earnestly parroting ChatGPTese like that was one of the creepiest things I've seen this year.

The blurb is still accurate -- the homepage has no weather, no news, no sponsored links, etc. Google the company does advertising, but the search engine page itself is still really minimalist. I think a lot of people forget (or never knew) how incredibly cluttered and ugly search engines used to be:

https://www.webdesignmuseum.org/uploaded/preview/excite-1996-preview.png

https://c8.alamy.com/comp/TRRJR0/lycos-is-a-search-engine-that-was-popular-in-the-1990s-TRRJR0.jpg

https://sm.mashable.com/mashable_me/image/default/uploads252fcard252fimage252f1530168252fe88d0a82-60be-4830-83_1u6w.png

What is this criticizing, exactly? The Google homepage is still incredibly spare.

Yeah, this post shows a tragic lack of familiarity with the concept of zoomies.

"How do you kill that which has no life?"