I'm fascinated by some of the comments in here, that the videos are "awful," "far from good," "embarrassing," etc.
This is technology that was unthinkable before just a couple of years ago. Photorealistic video generated by typing out a description of what you want to see is something from a sci-fi movie. I must be dreaming, this can't be real.
I took a screenshot in the middle of a video, and it's indistinguishable from a photograph: https://imgur.com/a/TyNbd7E
I don't know, maybe I'm just old. I remember hacking the hidden iframe trick to make ajax-style file uploads work in IE7. The advancements I've watched unfold in the past couple of years are still hard to wrap my head around.
No one is criticizing the photorealistic video generation. Yes, AI has gotten very good at that, I don't think anyone is disputing that.
At some point you have to move past your astonishment at the technical achievement alone and judge the result for what it is, on its own merits, as if a human had made it. Especially when the goal of using generative AI is to remove as many human creators and as much human effort from the creative process as possible, and to have as much "art" be as fully AI generated as possible.
People are criticizing these videos because they aren't good. As "art." Which is a problem if this is what all art is supposed to be become.
> judge the result for what it is, on its own merits, as if a human had made it.
I don't think anyone was proposing this as an actual professional music video, right? It's a tech demo that used all of $25 of budget. So why are you judging it "on its own merits, as if a human had made it"?
Most of the "professional" AI generated content I've seen isn't much better than this. It all lacks in the same dimensions, it's all the same sort of mediocre and uncanny. Claiming "it's a tech demo, actually quality doesn't matter" isn't going to cut it when this technology is already replacing actual human professional work.
Besides, several people in this thread have already decided that these tech demos are better than human created music videos. In fact, many AI "artists" already claim generative AI has surpassed humans in every relevant metric and they judge it on the same merits as I do just coming to the opposite conclusion. They certainly aren't waiting until it's "good" to call it art.
So it's entirely valid to point out that this technology is actually bad at doing the thing it's supposed to do.
>I was on an airplane and there was high speed internet on the airplane… it’s fast, I’m watching YouTube clips, I’m on an airplane. It breaks down, they apologize, and the guy next to me is like “This is bullshit..."
>Like how quickly the world owes him something he knew existed only 10 seconds ago?
>Flying is the worst one, people come back from flights, they tell you their story and it’s a horror story. They act like their flight was like a cattle car in the 40s in Germany.
>"It was the worst day of my life! First of all we didn’t board for 20 minutes. And then we got on the plane and they made us sit there on the runway for 40 minutes. We had to sit there."
>Oh really, what happened next, did you fly through the air incredibly, like a bird? Did you partake in the miracle of human flight, you non-contributing zero?
>You’re flying! It’s amazing! Everybody on every plane should be constantly going ‘oh my God, wow!’
>You’re sitting in a chair in the sky
>….Here’s the thing, people say there are delays on flights. Delays, really? New York to California in 5 hours. That used to take 30 years. And a bunch of you would die on the way there. And have a baby. You’d be a whole different group of people by the time you got there. Now you watch a movie and you take a shit and you’re home.
I'm guessing you've never experienced the enormous pressure of needing to find a job to buy food and clothes for your family. That's good, I'm glad that you don't know that feeling. But if you did, you'd know how easy it could be for a person to start feeling more and more desperate for any kind of lifeline.
There is still no chance that my first reaction to a random stranger asking me to do work for them is "sure", without building some sort of connection. Granted, that means I could still get phished via a coding exercise, but it would require a bit more effort on the attacker's part.
>The goal should be to reach as many people, of course, but also to ensure that the method and medium of communication is in the interest of the public at large.
Who decides what communication is in the interest of the public at large? The Trump administration?
You appear to have posted a bit of a loaded question here, apologies if I'm misinterpreting your comment. It is, of course, the public that should decide what communication is of public interest, at least in a democracy operating optimally.
I suppose the answer, if we're serious about it, is somewhat more nuanced.
To begin, public administrations should not get to unilaterally define "the public interest" in their communication, nor should private platforms for that matter. Assuming we're still talking about a democracy, the decision-making should be democratically via a combination of law + rights + accountable institutions + public scrutiny, with implementation constraints that maximise reach, accessibility, auditability, and independence from private gatekeepers. The last bit is rather relevant, because the private sector's interests and the citizen's interests are nearly always at odds in any modern society, hence the state's roles as rule-setter (via democratic processes) and arbiter. Happy to get into further detail regarding the actual processes involved, if you're genuinely interested.
That aside - there are two separate problems that often get conflated when we talk about these platforms:
- one is reach: people are on Twitter, LinkedIn, Instagram, so publishing there increases distribution; public institutions should be interested in reaching as many citizens as possible with their comms;
- the other one is dependency: if those become the primary or exclusive channels, the state's relationship with citizens becomes contingent on private moderation, ranking algorithms, account lockouts, paywalls, data extraction, and opaque rule changes. That is entirely and dangerously misaligned with democratic accountability.
A potential middle position could be ti use commercial social platforms as secondary distribution instead of the authoritative channel, which in reality is often the case. However, due to the way societies work and how individuals operate within them, the public won't actually come across the information until it's distributed on the most popular platforms. Which is why some argue that they should be treated as public utilities since dominant communications infrastructure has quasi-public function (rest assured, I won't open that can of worms right now).
Politics is messy in practice, as all balancing acts are - a normal price to pay for any democratic society, I'd say. Mix that with technology, social psychology and philosophies of liberty, rights, and wellbeing, and you have a proper head-scratcher on your hands. We've already done a lot to balance these, for sure, but we're not there yet and it's a dynamic, developing field that presents new challenges.
I remember in some countries there's an official government newspaper. Laws reference publishing things in this paper (e.g. tax rate changes, radio frequency allocations) and the law is that you must follow it once it's published.
In practice the information is disseminated through many channels once it's released in the official newspaper. Mass media reports on anything widely relevant, niche media reports on things nichely relevant, and there's direct communication with anyone directly affected (recipient of a radio frequency allocation) so nobody really subscribes to the official government newspaper, but it's there and if there was a breakdown of communication systems that would be the last resort to ensure you are getting government updates.
>The positives you experienced are very possible for a homeschooled student as well, and this seems to be a common boogieman.
How do you do that? Seems like it would be impossible to replicate the experience of learning to navigate daily social interactions in a mixed group of people, especially when it comes to dealing with conflict.
Easy - homeschooling may include but does not require "in the home" any more than "homework" is required to be done in your house.
I was homeschooled and have homeschooled my three kids. Never has that meant "only at home and only with my family". My kids have been in co-op classes, taken classes from Art or Technical instruction centers (piano lessons, voice classes, programming, robotics), enrolled in community classes via private institutions and the local JC (cooking classes, performing arts) and been enrolled in independent study charter public schools which have some in-person classes. And in high school they start taking in-person JC courses.
There is lots of regular exposure to a variety of other people in all of that!
> especially when it comes to dealing with conflict.
What makes you think school is a good environment for that? Kids can be very cruel to each other with often the most societally maladapted dominating for reasons that have no bearing in real life.
But users like me who hate shorts so much that they want to disable them in the app aren't addicted to shorts because we refuse to open them. And there's no risk of me going to Tiktok or reels because I hate short-form video.
Why would a user who hates shorts so much that they want to disable them in the app be sharing links to shorts with their friends?
If a paying user want to disable shorts, wouldn't allowing that ability make it more likely they will continue to pay?
The reason I started paying for Youtube premium was to turn off the ads. I hate YT shorts and I get annoyed when I accidentally open one. If YT continues to shove shorts down our throats, I'll probably cancel my subscription because I hate shorts that much.
> Why would a user who hates shorts so much that they want to disable them in the app be sharing links to shorts with their friends?
Because the user thinsk it's a funny penguin and that their friend will laugh. The reality is that for almost all users, the demonstrated and disturbing reality is that they will engage with what you put in front of them if you can tune it right. They may wish you didn't do so, and may idly lament to people about how much they resent you for not giving them more control, but they still engage, and in cases like yours, still subscribe. They're that attached (addicted) and therefore that valuable.
> If YT continues to shove shorts down our throats, I'll probably cancel my subscription because I hate shorts that much.
What modern online media companies learned is that they really don't have to care about that. Individually, you and your subsription don't matter to them at all, and most people just don't get indignant enough to storm off over stuff like that as long you you put the right funny penguins and half-naked women in front of them, so it all works at scale regardless.
And if you were to cancel your subscription, are you ready to go so far as to give up the platform entirely, or would you just fallback to being an ad target who's demonstrated all the appealing targeting characteristics you already have, while still being fed shorts?
That seems extremely unlikely given the significant global economic impact of tariffs, and the comparatively microscopic effect of transgender athletic participation in the United States.
For example, when Kentucky passed their trans sports ban in 2022, there was a grand total of one (1) transgender high school athlete in the state[0].
I mean, it's almost laughable to suggest that any of those things are even near comparable to the potential long-term impact of historically unprecedented tariffs being thoughtlessly tossed around on a whim.
They are related in the sense that stuff like violent crime by ethnic gangs and transgender affirmation of children easily rile up people to Trump or others who will bluntly oppose it, and then you end up with poorly designed tariffs.
It was largely a failure of anti tariff other candidates to capture these other 'easy wins' needed to get to the point of implementing sane economic policy.
I believe the claim was, with Trump in power there will now be a proper investigation of voter fraud in the 2020 election.
I have doubts. First, as someone else pointed out, there already has been lots of investigation in Republican-run states, and they turned up very little. Second, Trump seems to have the ego-driven need to deny that he lost. He seems to have the tendency to hire people who tell him what he wants to hear, and to fire those that don't. That's not likely to drive an impartial investigation.
In France, vote by mail was used between 1946 and 1975. It was stopped due to too much fraud (even if less than 2% of the population used it).
Just the fact requiring a valid ID is a contentious idea in the US is baffling when seen from here. I think more people in the US should check how elections are done in other countries (and not just 3rd world ones).
>>It is unlikely that there has been a significant increase in electoral malpractice since the introduction of postal voting on demand in 2000; available figures suggest that 32 convictions were made from 1994–99. In both periods, the offences arose almost exclusively from local elections, and related to a tiny proportion of all elections contested.
Again, the facts are clear, but that doesn't stop the baseless fearmongering. A direct quote from your source:
>>There is no evidence to date suggesting that electoral malpractice has occurred as a result of pilots of various forms of electronic voting. However, serious questions about the security of electronic voting from organised fraud remain unanswered.
It flatly concedes that no evidence of voter fraud from electronic voting exists, but then somehow concludes that serious questions remain unanswered. Simply absurd.
None of this changes the fact that mail-in voting in the United States is objectively and measurably secure, and that instances of fraud are so miniscule that claims of it having an impact on election outcomes are provably false.
What would your response be to someone who, hypothetically, might view this sentiment as some unsettling combination of elitist and vaguely authoritarian and then decides that maybe you shouldn't be voting? Would you be bothered by that?
This is technology that was unthinkable before just a couple of years ago. Photorealistic video generated by typing out a description of what you want to see is something from a sci-fi movie. I must be dreaming, this can't be real.
I took a screenshot in the middle of a video, and it's indistinguishable from a photograph: https://imgur.com/a/TyNbd7E
I don't know, maybe I'm just old. I remember hacking the hidden iframe trick to make ajax-style file uploads work in IE7. The advancements I've watched unfold in the past couple of years are still hard to wrap my head around.
reply