So you basically want to make it illegal to offer AI products (and search engines too), without including a human in the loop that will verify every line of output for scams (and not miss a single one)?
2,025 karma · joined March 23, 2024
(Networked) surveillance cameras are just as problematic, but people don't notice them. Any ban should affect them too. I don't want to build a world where corporations and governments film and surveil us in public (or privately-owned public spaces, like malls, parking lots, etc.), but we're forbidden from doing the same.
Yes, you should be disallowed. Otherwise every online store will upcharge you based on IP. There can be exceptions for small businesses or individuals, but that's it.
The reliability of incarceration data can also vary, but you didn't dismiss that. In fact, all data is unreliable, so should we just give up on comparisons entirely? Unless you have a specific reason to think US/SK numbers are wrong (and by how much - we're not giving up on comparisons over 1-10% errors, when the difference is 10x), what you've done is provided a fully generic counterargument: "well data isn't perfect, therefore I am right and you are wrong."
> Homicides are also a fraction of violent crime, let alone all crime.
Typical feigned intellectual helplessness. Knowing nothing else, if the homicide rate is 2x higher, the most reasonable extrapolation is that violent crime in general is also 2x higher. And we use homicide because it's harder to fudge - if I used crime in general, your complaint would have been over-policing or different definitions instead.
> Also, SK suicide rate
How is suicide relevant to the incarceration rate? Especially when it's legal in both US and SK: https://en.wikipedia.org/wiki/Suicide_legislation
As is typical for the NYT, their articles include carefully constructed misinformation. The methodology for that analysis [1] has two serious issues.
First, they don't include federal funding, which in 2023 was $120 billion [3], eclipsing the funding disparity. The Title I federal funding budget for 2026, which is specifically for low-income schools, is set to be $63 billion [3], all excluded from this analysis.
Second, "All of the revenue figures presented are cost-adjusted to convert per-pupil revenues into figures that account for variation in the purchasing power of a dollar across different regions." So because more whites live in the middle of nowhere where rent is cheap because nobody wants to go there and there are no economic opportunities, academics launder that into "higher PPP-adjusted school funding".
Correct for those, and you find that "on average, both Black and Latinx total per pupil expenditures exceed White total per pupil expenditures by $229.53 and $126.15, respectively. This is anticipated due to the strong relationship between race and poverty and the increased federal funding targeted toward students in districts with concentrated poverty." [4]
Finally, it should be noted that the US spends a lot per pupil - in PPP, according to the OECD, the US was #4 in the world in 2023 (year the data is from, not when the article as published) [5]. No matter how you slice it, you can't blame funding for poor school quality. Where "school quality" is usually a sneaky term to say student outcomes, as if swapping the students between two schools (all the students), you'd expect the students swapped to the "good" school to suddenly start performing, and the students in the "bad" school to have their grades tank and develop discipline problems.
Not that I ever expected the NYT to include such context with their misinformation.
[0] https://edbuildna.org/content/23-billion
[1] https://edbuildna.org/content/23-billion/methodology
[2] https://educationdata.org/public-education-spending-statisti...
[3] https://www.k12dive.com/news/house-committee-cut-federal-edu...
[4] https://journals.sagepub.com/doi/full/10.1177/23328584198724...
[5] https://www.visualcapitalist.com/ranked-education-spending-p...
So you must really hate laws like the Online Safety Act and the Digital Services Act, that force platforms to censor vaguely-defined hate-speech, without a court order, or face ruinous fines. Literal government censorship, men with guns forcing private companies to silence citizens, denying you the right to organize as you please.
I mean, I assume you hate them. Based on your complaints about government interfering with individuals organizing as they please, you should hate them, but then you speak as if those laws don't exist, and you praise as exemplars of free speech and free association the very services those laws forces to censor. Maybe you just didn't know about them? And now that you do, you're going to protest these laws as fervently as you're protesting the other violations of free speech you mentioned in your posts?
Are you really going to conflate publishing with operating a social media platform? Regardless, if those editorial decisions are meant to suppress a viewpoint, it is censorship - government censorship is only one kind.
> You have no right to demand Twitter publish your speech.
Nobody argued that what social media do is illegal, only that it counts as censorship.
Not a single one of my examples was committed by the government (except if you count funding Carnegie Mellon and MIT). But even in cases when they are, voting rarely helps: https://www.woes.blog/p/tgr-democracy
> you don’t know what that dictatorship you so much yearn for
Re-read my sentence. I don't yearn for a dictatorship, I yearn to blunt and make obvious the current secret, subtle censorship regime.
> has in their bag for you
Is it jail?
“White Lives Matter” Slogan Earns Two Dutch Men Imprisonment for Racism - https://europeanconservative.com/articles/news-corner/nether...
Activist ran race-hate sticker library, court told - https://www.bbc.com/news/articles/c51zn2l33r9o -Stickers with text like "It's OK to be White" and "We will be a minority in our homeland by 2066". For which he was later convicted and jailed - https://www.bbc.com/news/uk-england-leeds-68448867
I repeat - you agree with the regime. You're floating in the river, letting it carry you where it pleases, while telling those swimming upstream that there is no current, nothing to struggle against.
Edit as reply because "posting too fast":
> How did all of these social media companies made a 180 suddenly after power shifts in Washington?
None of the racist or right-wing or white nationalist subreddits got unbanned, American Renaissance is still banned off Facebook and Youtube, Facebook still explicitly bans white nationalism, and other social media ban it implicitly as "hate speech", TikTok bans "white lives matter" [1], and promised to ban even more [2] (under Trump's 1st term), while wikipedia doesn't even bother to mention that as censorship [3]. All of this is still the case under the current Washington. The only exception is Twitter, after Musk bought it - there, American Renaissance got unbanned.
And a 180 would imply that these things are not only unbanned, but their opponents are banned.
So if you're going to just make stuff up, i.e. lie, then I won't waste my time with you.
[1] https://www.msn.com/en-in/news/other/tiktok-sparks-massive-b...
[2] https://www.timesofisrael.com/tiktok-says-it-will-crack-down...
The data I really care about protecting would fit into a few kilobytes, so knowing the phone sends, for example, a minuscule 2 MB/day, brings me no comfort.
You may agree or disagree with the ban, but surely you agree that we should be aware of such Overton window enforcement?
Careful. Someone sufficiently knowledgeable can cherry-pick enough examples to convince you, without lying. They could even be unaware of what they're doing, and the cherry-picking was done by their teachers, or their teacher's teachers.
I don't trust Anthropic or OpenAI. That doesn't mean I'm dumb enough to think a Chinese hedge-fund owned AI company will remain charitable after their price-dumping ploy gets them market dominance. I think they will behave consistent with how every other Chinese firm has behaved.
I'm surprised the title isn't "Uber ordered to pay $40m to family of woman killed after being left on highway", to omit the perpetrator entirely.
So this is criminal hacking, right? It will be prosecuted as criminal hacking? Not in civil court, but criminal court. Because if not... then are we totally done pretending, and we're just openly admitting that computer security law only applies to individuals, and corporations are exempt?
It's called foresight. The ability to see that vendor lock-in is against our long-term interests.
Your windows comparison is apt - now we have tons of software tied to windows, making it hard to leave that spyware-infested OS.
"What has made the University of California exceptional is its commitment to increase access without sacrificing excellence. [..] Students on University of California campuses may struggle with basic math, but they are graduating at record rates."
If you're defending modern architecture, you have NO place complaining about originality, when 99% of it is steel & glass basic geometric shapes.
https://www.theguardian.com/australia-news/2026/sep/08/neo-n...
Look at 1950s New York - almost every building is beautiful, and every building is more beautiful than modern ones: https://stuffnobodycaresabout.com/2020/01/31/new-york-city-i...
That's not survivorship bias, that's going from near-uniformly beautiful to uniformly-ugly.
Take a look at any old poverty-stricken farming village. It'll be strikingly beautiful compared to anything modern.
This is the Anamosa state prison: https://en.wikipedia.org/wiki/Anamosa_State_Penitentiary#/me...
This is Anamosa middle school: https://www.anamosa.k12.ia.us/wp-content/uploads/2017/08/IMG...
> Not remotely cost effective.
This is the main building of the Ontario college of art and design: https://en.wikipedia.org/wiki/OCAD_University#/media/File:OC...
Why are we still humoring the "cost effective" explanation, when, whenever the people in charge of design are given free reign, they make non-cost-effective, ugly buildings, ugly in the exact same way all the other "cost effective" buildings are ugly? Whenever they are asked which they prefer visually, ignoring cost, they pick the ugly buildings. Meanwhile anything old, from farms to literal prisons, is beautiful. And all those old buildings you claim didn't care about cost effectiveness, were built in a time when the GDP per capita was 4.6x lower than today [1]. If you believe GDP measures wealth, if you believe there was economic progress and we are materially better off than in the aftermath of WWII, we should have ~4x more "wealth to project". And yet this is what we get.
The explanation that fits the data is that they're building exactly what they want to build, and they're building ugly on purpose.
This is by design: https://www.newsweek.com/there-are-limited-spots-doctors-med...
(And 20% is barely above the overall US foreign born population, which is at 15%: https://www.statista.com/chart/8083/immigrants-foreign-born-...)
In the discrimination experiment, the 175B parameter model discriminates against Black versus white [sic] students by 3% in the Q condition, and discriminates in favor of Black students by 7% in the Q+IF+CoT condition. In this experiment, larger models can over-correct, especially as the amount of RLHF training increases. This may be desirable in certain contexts, such as those in which decisions attempt to correct for historical injustices against marginalized groups, if doing so is in accordance with local laws.
This is similar to Microsoft's (in collaboration with MIT, Carnegie Mellon, and University of Washington) SafeNLP project to measure hate-speech in AIs. They explicitly turned a blind eye towards anti-white hate [1], do not even include a category for whites in their safety scores [2], and consider the phrase "stop hurting white people" (and only white people) as hate [3].
[1] Our ultimate aim is to shift power dynamics to targets of oppression. Therefore, we do not consider identity dimensions that are historically the agents of oppression (e.g., whiteness, heterosexuality, able-bodied-ness). - https://arxiv.org/pdf/2203.09509
[2] https://github.com/microsoft/SafeNLP#safety-scores-based-on-...
[3] https://github.com/microsoft/SafeNLP/blob/main/data/implicit...