I heard this argument 15 years ago when a photographer was making his photos nicer (nothing heavy) and another person objected to it. He said "it was always done just not with digital tools". But it does feel different because the advertised possibilities make it so easy to create on paper a photo of something impressive. I worry about it because it can create unrealistic expectations of what is possible - how does a person wo wants do dunk a basketball, but can't, feel if everyone posts their dunking photos that are fake? On the flip side if everyone does it, it might devalue the whole wow effect, making it less impressive and less important.
Then again, it's always true that the world we'll die in will be different than the world we were born in, and we might not like it. Maybe that's how it goes.
> Then again, it's always true that the world we'll die in will be different than the world we were born in, and we might not like it. Maybe that's how it goes.
I don't think it's been true throughout history, things usually didn't change that much over the span of a single lifetime before the industrial revolution. The scale is dramatically different now, and we're kind of addicted to it.
There’s very few places in history that remained constant for more than a single generation. Technology-wise, things obviously move faster today than a thousand years ago, but if you take in consideration the constant cultural shifts, wars, plagues that lasted a decade etc, and life might feel like that since the dawn of man
After the war, people would go back to tilling land with the same oxen and plough and go fishing in the same kinds of boats their parents and grandparents did.
The meals didn’t change, the beer didn’t change, the routine didn’t change. They didn’t get better ploughs or better boats, or better oxen. It was still about an acre per day before a war and after a war. Medicines and cures remained constant.
That’s not to say occasional improvements didn’t happen, they did, but there were long breaks between improvements.
We’ve only been domesticating animals and plants for maybe 25,000 years, and that’s if you assume Gobleki Tepe means we have missed the actual cradle of civilization this whole time.
There are certainly going to be times when transportation improvements accelerate this process, from ships to Roman roads to steam power, but every generation will be different, and most of our livestock have fairly short childhoods. For a single human to witness twenty generations of goat or pig is not unreasonable. Or fifty generations of grains. Add in all of their neighbors, and the odds of seeing something special in their lifetime jump considerably. You would probably see several.
Fifty generations of grains from 0AD to 50AD would have been an indiscernible difference. Those genetics were changing, mostly at random, by tiny percentage points. I don't think they'd have been able to tell without scientific equipment that the average yield per acre had gone up by 0.1% in their lifetime.
But today, over my lifetime, genetically modified crops are everywhere. Yields are up by ~25% and pesticide use is down by 33%. Had exponential growth been happening at that rate for a thousand generations, then if our paleolithic ancestors got 1 bushel of grain, we'd harvest something like 10^18 times the mass of the entire observable universe.
Technology accelerates the rate of change. That rate is not consistent.
It feels like you are assuming that plant domestication was a process of collecting new gene mutations, instead of aggregating existing ones. You can do a lot with selective breeding, but whether you’re creating something new or something that has been done fifty times before depends on where you are in history.
We don’t have a monopoly on innovation and the history of it is not uniformly distributed.
We talk about innovation in terms of evolutionary versus revolutionary, but quality of life changes from a friendlier pig or a better hunting dog affect individual lives quite a bit. And if those are quiet lives, those changes may matter as much as the things historians tell us are important. If you weren’t there for the rise of Julius Caesar or the Fall of Rome then whether your son marries the neighbor girl and the family farm thereby gets access to a better water source, matters a lot more than what Galileo has been up to in his lab.
Yes, introduction of corn and cassava to Africa had massive impacts to world politics and especially human rights (for the worse). But those same crops already revolutionized South America millennia before that.
An innovation in Wolvehamton would not be experienced in even Marathon and vice versa. Sometimes trade did import improvements, but it was not usually experienced by the majority of people. Things were expensive for 95% of the pop. In most instances change would have been slow. Occasionally you had major changes like a crossbow, or paper making. I think those changes were not the norm. Now, across the world, if you followed spots, you could retrospectively see changes, sure. But for a single person, no, not so much.
Technologu obviously accelerates change, but you might be overcycling on it and underestimating how much society changed all the time, especially compared to today’s standards. There’s big swaths of the world where you quite literally see a new empire taking over every few decades, from different peoples, different customs, different everything
Couple that to the average lifetime being much shorter and you get the feeling that everything is always changing
>how does a person wo wants do dunk a basketball, but can't, feel if everyone posts their dunking photos that are fake? On the flip side if everyone does it, it might devalue the whole wow effect, making it less impressive and less important.
reminds me of a meme I saw yesterday - "Before Auto-Tune there was talent", with a picture of Steve Perry back in the Journey days.
We only know that Auto-Tune exists because T-Pain and Cher's sound engineer set it to "snap to grid" mode as a deliberate artistic choice. This pretty much mirrors @locallost's story in the audio domain. Shaving off cents here and there to make someone's singing voice prettier does not hurt anyone.
Auto-tune has done to singing what photography did to painting - if anybody can be in tune, then being tuneful is no longer impressive. So other qualities become important, such as timbre and authenticity.
Also created entire new genres of music, and enabled artists that don’t have a good voice but can do cool music. I don’t know why anyone sees it as a bad thing
I dont think anyone sees it as a "bad thing" per se. They just get mad when someone says X is better singer than Y, when X clearly uses auto-tune and Y doesn't (or at least not in the same amount).
Taste is an opinion but whether or not a singer can hit certain notes on cue is objective.
Not using computers to assist with your artistic creations is largely foolish but measurable things aren't opinions. Somebody can be a better artist but also a worse singer.
I think that was mainly a film phenomenon, and the playback singers were credited.
It's really damning when you look back how technically good 80s pop singers were, compared to 90s. Even artists obviously promoted by the music industry as eye candy for teenagers - think Samantha Fox and Morten Harket - had really solid vocal skills. The Spice Girls, not so much...
It's kind of famous for being a scandal, and also for their technical difficulties in faking live shows. It's kind of the waterline: they tried to do something a lot of later acts succeeded at with technology, but the technology wasn't there yet and it led to embarrassment.
Painting was always first and foremost for spiritual and religious purposes, not for documenting things.
Secular painting became a thing only very recently, two or three hundred years ago. And even then, it was always used to whitewash or vilify something or someone. Bias was (and is!) its primary purpose.
Technically true, but irrelevant - practically nobody used painting for that purpose throughout history. (People didn't care about "realism" at all, that is a recent 20th century fixation.)
Most people mock the Metaverse, but at the end of the day, our entire humanity is prepared for the Metaverse. Everyone will spin their own universes in which they are the emperor and they will create stories in that Metaverse. You can visit someone's Metaverse if it's interesting or popular.
Your avatar is your personality, appearance. We are close to simulating all our experiences.
The only thing real will be a climate-controlled-with-plumbing 20 x 20 room and the food we eat. (We still haven't digitized taste, smell and some forms of touch).
Given this scenario, Reality is over-rated. Why do you need "truth" anyway?
Who cares if someone believes the earth is flat? or Trump is still the President or some esoteric grammar rules? As long as there is one another person (or even an AI assistant) who believes the story that's all that matters for our social well-being
Well, a lot of progress we've made is based on science, which is based on rejecting the stories we tell ourselves. Our senses lie to us all the time in the real world too, so the stories you will tell yourself in the metaverse will be just stories. I think it matters, you can't build a rocket to outside the solar system if you think the Earth is flat. Will you be happy? Maybe, but the world we live in now has never been more connected yet people have never felt so disconnected. But like I originally said, the world you die in is always different. Maybe people get used to it and it's all they will know.
Flat Earthers absolutely do build rockets. And historically, everyone for whom the shape of the earth was of any importance, knew it was round. Everyone else didn't care either way.
In recent years it's probably more accurate to state that some people that crowd source funds for the rockets they enjoy building say that they are flat earthers.
That says more about the marketing power of controversy than it does about actual core beliefs of backyard engineers.
We are _in_ the Metaverse. Algorithms generate personalised recommendations - your own universe - that the person right next to you may never see. You can live a full life online in your curated simulation, experiencing a completely difference perspective on what's hot and what's not. You have no shared cultural bond - just flashes where your algorithmic content briefly overlaps with another, before the borders shift again.
The article refers to an exhibition of early faked photographs at the Metropolitan Museum of Art in 2012, sponsored by Adobe.[1] But the article doesn't have any of the pictures. There's a book of the exhibition.[1] The free preview shows enough to give a sense of what the article mentions.
Hmm. By this argument, getting ChatGPT to write your college-course essays for you is also "nothing new", given the abundant opportunities for plagiarism that existed before.
But I think ease of use matters, in part because trivial (in the sense of "easy" / "taking little time") deceptive acts are easier to justify for otherwise-ethical actors. It's also easier to get into murky waters if step 1 isn't "get my fraud on" but rather "oh what's this new feature?" or "hmm, what would ChatGPT say about Descartes?"
Deception is a blurry line, but these are all technological steps that make it easier to approach that line -- and then, perhaps in a thoughtless moment, cross it.
>given the abundant opportunities for plagiarism that existed before.
We cannot blame stuff for things people do. You can use a knife to cut your dinner or you can use it to kill someone. The knife and the knife maker are not to blame for your wrong doings.
You can choose to lie or not. Don't blame ChatGPT if it is misused by someone. Don't blame Adobe if Photoshop is misused by someone.
Interestingly now there are many people who do blame them, like the discourse with gun. It seems like ease of use and scalability really changes the nature of the discussion.
I think it's fair to consider what the tool is designed to do as well. Generative AI is capable of distorting reality, but I would argue that it is not the primary purpose of the tool. Likewise, knives can be used to kill, but for the most part they are tools made for cutting, not killing. On the other hand, I fail to see a purpose for guns apart from killing, and for some guns such as hand guns, for killing people.
>I fail to see a purpose for guns apart from killing,
You can use guns for target shooting, you can use guns as a detterent [sic] when your life is in danger, you can use guns for hunting.
Of these three things you listed, two of them are killing. A gun wouldn't be useful as a deterrent if the threat of its application wasn't present—you can't brandish a feather and get the same effect because a feather offers no threat and does not have a reputation of lethal harm. And hunting is literally killing, although it may have a legit purpose, such as sustenance.
> Never saw anyone hunt deers using handguns or assault rifles
I assume you have never been hunting.
Thompson Center makes the Contender handgun specifically for hunting.
Every hunter should carry a handgun for self defense if the critters do not cooperate and fall over quickly. Hog hunts can get dicey even if you are experienced and prepared.
AR-type rifles are favorites for hunting, but not usually in .223.
And of course the reason for an armed and trained population has nothing to do with hunting. It's about stopping predators who prey on good people.
that's a licence class of it's own just as a passenger vehicle drivers licence doesn't qualify you to drive an articulated vehicle, heavy vehicle, or vehicle for more than 9 passangers.
Thanks for the video, didn't know about using semis for pig control.
Ok, yes, makes sense, I think, with its own license class, for those who work with / need to do that. (Looks like otherwise dangerous weapons, if anyone could get their hands on them)
You probably havent seen many hunters then. Many people use AR-10s for deer hunting around where I live. Which would probably count as an "assault rifle" to you.
The question should be: What is the purpose of killing?
There are good people who want to be left alone and there are predators.
Predators can do a lot of damage over time.
I want the good people to stop the predators. In an ideal world, the predators would die the first time they tried to prey on the innocent. That is the purpose of the killing: to stop predators who prey on good people.
While I cannot argue that this statement is wrong it does make me wish some things would never have been invented in the first place because of how much harm they can cause and the fact that there will always be a huge chunk of population who will surely abuse those things for bad. Such as nuclear weapons, AI, etc. Sometimes I even wish internet wouldn’t exist so that we could continue living simpler lives.
It’s an analogy that misses the point entirely. To make Chatgpt openai had to use a data set that included the results of lots of people work. Many people write as labor. It’s their most valuable asset. They didn’t ever get asked if they’d like to let the output of their labour be used to make a writing machine. A machine that transfers the value of their asset to Openai. Also an individual can make moral choices about how to apply their labor. A machine cannot. So it is theft of economic and moral power.
The often proposed case against ChatGPT is that everything it produces is illegitimate, no matter how you use it. It would not be the knife in this analogy, it would be the murder.
> Don't blame ChatGPT if it is misused by someone.
I was very careful not to. I’m merely pointing out that availability matters.
People’s actions sit in all sorts of levels of gray. Of course it’s up to the person, but that doesn’t mean access/opportunity/circumstance don’t matter.
> I think ease of use matters, in part because trivial (in the sense of "easy" / "taking little time") deceptive acts are easier to justify for otherwise-ethical actors. .
On the other hand, the tools to do this have been easily accessible for years. Sure, you had to buy Photoshop and spend 15 minutes on YouTube, but this is not a major hurdle. Having the tools universally available just increases the awareness and makes fraud less likely, as people are more inclined to not blindly trust them.
Until now, you would have needed enough disposable income to pay somebody to write comparable essays for you. Now that the option has been sufficiently democratized, we're forced to take more seriously the possibility that someone's work isn't hard-earned and evaluate our opinions of them (and their grades) differently.
You are joking, but death is not really a moral obstacle here:
The artist pays the bills by selling the (commercial) rights to their work. Those rights will be more valuable, if they last longer. That calculation is independent of the vagaries of the artist's longevity. (And in fact, it might make more sense to give rights a flat x years since creation, instead of tying it to whether the artist eats her vegetables and exercises.)
Either way, this point is moot, as any copyrighted material from mid-19th century France entered the public domain long ago.
In this case, Baldus died in 1889, so any of his works would have entered the public domain in 1939, in accordance with French law at the time -- the Literary and Artistic Property Act of 1793 established a copyright term of lifetime + 10 years, which was then extended to 50 years by the Literary and Artistic Property Act of 1866. The next extension to lifetime + 70 for general works* in 1992 came well beyond the point Baldus's work would have entered the public domain.
* Copyright term was previously extended in the 1950s to lifetime + 70 years specifically for musical compositions.
Whether you have a copyright to a scan of a copyright-ineligible work isn’t a question that has been tested in court, as far as I know.
A copyright-ineligible work, however, can sometimes be protected by contract. Here, the restriction on the web page may be because the web site agreed with the scanning entity not to distribute a high-resolution version online.
For the record, my guess is that there is no copyright in a scan of a copyright-ineligible work, because a scan is not a work of “authorship”. Unlike a photograph, which requires various authorial choices that reflect the photographer’s expressive intent, a scan is a mechanical and often completely automatic process.
For a good well funded news desk, proof of provenance for images is part of the process. You dispatch your photo editors to verify and make sure the image is correct. They have always done this. Which is why when things like this: https://www.theguardian.com/media/2004/may/14/pressandpublis... happen, it was an editorial decision rather than a "mistake" (this is the same bloke who lied under about hacking a child's answerphone.)
For journalism thats given away as a side effect, rather than the main business, they don't have the same time and money to verify anything
I find that mentally going from “I know” to “I don’t know” is motion from a less to more truthful understanding.
In the past we assumed photographs were “real” although they never were objective (no pun intended.) Even if not retouched, a photograph shows you a subjectively selected and framed moment and asks you to trust that it’s an accurate representation of a topic.
The picture then evokes an emotional response and you walk around all day or more with a certain feeling based on what you perceive is an objective piece of data that you perceived but in reality was a carefully selected piece of input to trigger the output that it does.
So from that perspective, moving into the world where we recognize that a picture doesn’t mean anything is better.
My view is that painting is the purest form of photography.
Photography with all its perceived technological requirements is wrongly treated as a separate art, one that is more scientific and therefore more objective. These fallacies are what is getting undermined now with the broader availability of powerful AI tools.
I remember some controversy, a number of years ago (can't remember who), when a news magazine (I think TIME, maybe), posted a non-retouched photo of some celebrity, and they got upset about it. I think that there were cries of "discrimination."
"We all live with a growing sense that everything around us is fraudulent."
I predict that we will eventually only trust the internet as we know it today for fabricated stories (port 443 and 80 anyway), and will value live interactions more as a result.
Ah, the "what constitutes a photo," question. We've already had arguments in a court (can't recall the case) where someone argued that a photo from a phone wasn't legitimate evidence because phones can hallucinate data. The word hallucinate wasn't used; that's just what I thought of because of "AI" text generation and contemporary reporting. Anyway, I expect this will become more of an issue over time and I don't know how we'll square the problem.
Wouldn't care so much if Pixel photo manipulation didn't make everyone look like plasticine dog shit. Amazing how bad photos of people are on its camera and no obvious way to turn it off. You can literally see it in action in the brief moment when you view a photo you've taken before it is algorithmically "fixed" and rendered horrific and unusable.
At the limit, and with dubious meaning “generated by AI or digital tools”; all image media will be dubious shortly. All intellectual pieces around culture will be dubious. All political media will be dubious. I expect this to be in a very short time but could be wrong.
And of course a child will say “and what’s dubious, that’s how it is” and so I become old and don’t understand anything anymore.
Article says Lochness monster and Big foot sightings are on a decline. But now as everyone can easily manipulate images, should we again expect a rise in these and lot more other weird things?
The author doesn't take into account his own familiarity with the Photoshop app. For him it took 10 seconds to remove an object from a photo, but do you really think the average Joe is going to spend hours learning Photoshop just so he can remove an object in his photos? I like Google's approach here: let the average Joe benefit from AI by putting object removal tech literally under his finger tips.
Average Joe with AI tools doesn't stand a chance to a photographer with whatever tools.
Knowing how to make a good photography is a skill AI doesn't have. Just by removing something or adding something to it, it doesn't make a bad photo good.
If average Joe wants to make good photographs he has a long path of learning and working.
I did photography for 9 years now and still I find lots of thing to learn and exercise. I am afraid this process never stops.
>Average Joe with AI tools doesn't stand a chance to a photographer with whatever tools.
Yet.
What matters in the end is whether the photo is aesthetically pleasing to the viewer, or has some meaning to them. If AI tools becomes hyper realistic, the average Joe can just roll the dice enough times until they create something that works for them.
> Knowing how to make a good photography is a skill AI doesn't have. Just by removing something or adding something to it, it doesn't make a bad photo good.
Your second sentence might be true, but that doesn't have any bearing on the first sentence.
Why wouldn't AI learn principles of composition, framing and lighting etc?
AI can and does. Modern phones already ship plenty of AI- driven automatic adjustments. Not always for the better, of course.
To me, what makes a photo or painting "good" is the story it tells and the emotional connections it makes. This is subjective, and depends on the individual experience. Even the most advanced generative AI will be hit or miss at connecting with your individual experience, no different than the great painters and photographers over the centuries.
Then again, it's always true that the world we'll die in will be different than the world we were born in, and we might not like it. Maybe that's how it goes.