When I see a lot of this voice replication I am reminded of something.
(Dune series spoiler)
For example, Duncan Idaho keeps being brought back with cloning tech, over and over, for centuries, because Leto II just likes him. Over and over, the same man keeps being put by the (nearly) immortal emperor's side...
I ask: are we stuck as a monoculture, not making new, interesting things, but simply replicating older ones?
The technology is novel and interesting, but culturally, aside from the concerns about annoying the people being replicated, it seems a bit stuck.
On the short term, yes. On the longer term (probably only a year or two at this rate) no because these models also allow you to generate entirely new voices. We'll likely see features like describing the voice you want with a prompt and getting several to choose from, remixing existing voices, combining voices and creating voices that are impossible for any human to have. The introduction of other forms of recording and reproduction didn't result in total monocultures, even music which gets a bad rap for all sounding the same only has this problem in the mainstream, scratch the surface and you'll find hundreds of weird new genres and styles.
We could see a future where every commercial is narrated by Attenborough or Freeman but indie creators will be pumping out vast quantities of wild interpretations and new applications of the human voice.
That's basically what I said? Mainstream content will generally play it safe and still sell but there will be ever larger numbers of independent creators pushing the envelope.
This reminds me of something I noticed: disproportionately, the classics in a given field (ie, the work agreed upon to be particularly important, influential, etc) are disproportionately the ones that were made shortly after a new medium or technique was invented. For example,
It's almost as if we default to tending towards imitation and recycling, and only each time a totally new technology changes what is possible do we really try new things -- but the well of newness is rapidly exhausted and so those artisans who are fortunate to be able to be the first to work with a new medium get to be the ones that create its classics.
The question is why? Does a medium combined with a culture totally determine the kind of work that can be made? Or is the influence of the original classics impossible to escape? In other words, is this situation path dependent?
That seems just very obviously untrue. Take an example, say literature: Are you saying the classics of literature were produced shortly after the invention of the pen/paper (clearly untrue)? Or maybe after the invention of the printing press? (Also clearly untrue).
It seems like a banal observation to say when there is a technological breakthrough there are new discoveries made. Like say astronomy, every time we invent a new method to see further/measure better we discover new stuff. And equivalently in art as new media and methods are discovered, artists take advantage of them to produce new work. It's obvious both that this is the case and why that should be so.
Otherwise, we are clearly going to evaluate works that have stood the test of time to be "classics" and have difficulty making an objective evaluation in many fields of how new works stack up relative to them. History winnows out the things which aren't that great and then you have a new generation of classics that "by pure coincidence" (not really) are closer to when the new medium or technique than your time of observation.
I'm not sure that's true. Look at the printing press. Sure, Shakespeare is a classic, but most of the rest of what we read as "classics" probably centers around the mid-1800s (at least in English-speaking countries)
Yes, that's exactly some of the time periods I was thinking of. The mid 1800s was a time when literacy was rapidly increasing and new serial magazines became popular. This new medium paved the way for the fame of authors like Dickens and Bronte, the sort you may be thinking of.
This only works if you compress the entire 19th century into a single point. Otherwise you have, eg Sense and Sensibility from 1811 vs Heart of Darkness from 1899.
If anything this claim merely shows the effect of “time compression” in which distant events feel like they happened around the same time.
Not to mention Frankenstein and Journey to the Center of the Earth, both of which are science fiction classics. And then, of course, you leave out the entirety of 20th century science fiction.
You're cherry-picking to establish your conclusion. Lots of classic authors from that period (George Eliot for example) didn't serialise their works in magazines.
Also the 19th century is not by any means considered generally superior to any other period in English literary history. Eg Jane Austen and Virginia Woolf and James Joyce are looking sternly in your direction from either side of that timeline for example.
Or just the earlier works have been forgotten. It tends to surprise people when I mention the 1939 Wizard of Oz movie (one of the first(?) to showcase color and still known for that because of the transition) was an adaptation of just the first book of an older series.
Generations move on.. All familiar things are eaten by time, so it's familiar things in media from youth that provides an anchor. The city you grew up on is gone except for a few landmarks. There are always artist artists who experiment but that's not mainstream taste
> I ask: are we stuck as a monoculture, not making new, interesting things, but simply replicating older ones?
I really don't understand how you go from this story to that question. There are two, new, interesting things here:
1. The dev has made something which narrates their life. That's new and interesting
2. David Attenborough actually is still producing new (and amazing) nature documentaries. Planet Earth III has just come out and is astounding https://www.bbc.co.uk/programmes/p0gjwxhv
Attempts at cloning are often the dirty secret behind much lauded creativity.
I like the stories of unique art/music that was created because someone was trying to clone something exactly... but despite their best effort, they just couldn't get it right and ended inventing something new that people loved anyway.
We just have to make sure we are pretty bad at the cloning!
everything in culture is based on previous work and sampling; it has nothing to do with cloning.
If you can't speak the language of your field you can't produce something original. you have to master the basics first aka emulate others aka learn the language before you can become creative.
> I ask: are we stuck as a monoculture, not making new, interesting things, but simply replicating older ones?
Probably, but I hope not (or at least not for long).
I think it's normal for people to want to retell stories (see the way we pass jokes around like viruses, or the oral tradition, or the million "re-imaginings" of classics). I also think that inclination has been weaponized and turned against us. Producers -- movie, music, or other -- push or are pushed to maximize gain while minimizing risk. And what feels better than nostalgia? What feels better than a story that doesn't challenge you, doesn't prompt any sort of introspection, or hell, any inspection at all?
I think it's a symptom of the push for profit over all else. I do think we ache for new things, but the ship of monoculture is a big one and takes a long time to steer. A lot of momentum is behind where we're at and it's going to take a lot to get away.
> I ask: are we stuck as a monoculture, not making new, interesting things, but simply replicating older ones?
I've been thinking about this a lot lately. It's such an interesting side effect of the streaming era that some of the biggest TV hits with teenagers are Friends and The Office. Not to knock either show but I'm fascinated by what it means for culture.
Content creators now need to compete with the best that's ever been created, not just the best that happens to be broadcast at the same time.
This will increase the inequality amongst general purpose evergreen content, the winners will be bigger than ever before and continue to be watched for decades - but getting into that winning bracket will be hard.
Conversely it's easier than ever to get views on perishable content, that which has short relevancy to current affairs or rests on a particular cultural trend. Live streaming too, by it's nature resembles the rivalry dynamics of broadcast television.
Good point, and it mirrors what happened with popular music. Growing up in the 80s, I was almost completely unaware of music from the 50s, and didn't care for what little I'd heard. But plenty of younger people today listen to even-older music from the 60s through the 90s without batting an eye.
We'll see the same thing happen with movies before long. A cynic might say that, for various reasons, the best ones have already been made.
I'm excited to see the first brand new content from an existing series. I really love the 80's cartoon Inspector Gadget. There are about 100 episodes and they are all pretty repetitive (so hopefully easy to train AI with). I would be enthralled if someone could create new episodes with:
- AI generated storyline
- Deepfake the original actors voices
- Deepfake the animation, which should(?) be easier (more realistic) than the video deepfakes.
This could be so cool, "hey AI, make an Inspector Gadget espisode where the dog gets lost in the forest and Uncle Gadget eats magic mushrooms he finds while trying to find him"
I'd actually find that quite disappointing. If you can generate practically infinite variations of a show using AI, it completely eliminates scarcity and makes the result meaningless.
It's not even unique to AI really. The whole genre of games that depend on procedural generation are similar. Who cares if there are a thousand planets in Starfield if they're all just the output of RNG? There's no actual handicraft involved to set one environment apart from another. There's more music on Spotify than anyone could listen to in a lifetime; it just makes it harder to find new stuff that is actually good or else you stick to what you know.
> There's no actual handicraft involved to set one environment apart from another
one of the reasons I don't like modern animation is the lack of effort in the actual animating. South Park looks like it was drawn by 5-year-olds who traced the outline of a salt shaker for the characters. Inb4 "it's supposed to be that way, it's all computer assisted and has been for 20 years", yes I know that. The point is there's no art involved in it. 80's and 90's stuff was hand drawn (AFAIK) and often outsourced to animators in Japan, and yeah it's primitive, but its literally handicraft. The only scarcity of shows like southpark is people with squeaky annoying voices. Even the writing is so repetitive and predictable. It was novel and funny in the 90's-early 2000's, but "kid saying adult thing in exuberant voice" is tired and old. Simpsons were doing it in 1989 ffs.
> one of the reasons I don't like modern animation is the lack of effort in the actual animating.
The new Ninja Turtles and Spider-Man movies certainly contradict that, as do plenty of animations from Japan such as anything by Satoshi Kon.
Yes, there’s a lot more garbage today but that because there’s a lot more of everything. The good stuff is going strong and there’s innovation still happening.
South Park is a bad example because it’s atypical in many ways. They do the whole episode in under a week, including the writing and animation, and the cast is basically the creators.
I believe it is a temporary state of things, thought a sad one. ML made making mockups and mashups effectively effortless, but they are boring and should fall out of public curiosity spot soon, I believe. At the time Hollywood lawyers figure out how to put synthetic clones of big stars into AAA-movies, people will be too tired with sequels, prequels, spin-offs and stuff like that.
I mean, there's a whole lot more to it than just that. Duncan is a key figure to The Golden Path Leto's breeding program. Without Duncan, the success of creating Siona is hollow. It's their nine daughters and thousand sons who are the human protection from prescience.
Mark Fisher didn't create this word, but he wrote a lot in the 2000s about our inability to imagine a new future after the failure of our previous visions.
Using electronic music - a "future-looking" genre - as an example, he wrote (second paragraph explains it more generally):
"...Twenty-first-century electronic music had failed to progress beyond what had been recorded in the twentieth century: practically anything produced in the 2000s could have been recorded in the
1990s. Electronic music had succumbed to its own inertia and retrospection. It was also clear that this was more than a moment in a familiar pattern, in which, as one genre wanes, another emerges to take its place at the leading edge of innovation. There was no leading edge of innovation any more. In music, as elsewhere in culture, we were living, in Franco Berardi’s suggestive phrase, after the future.
...
What haunts the digital cul-de-sacs of the twenty-first century is not so much the past as all the lost futures that the twentieth century taught us to anticipate. The futures that have been lost were more than a matter of musical style. More broadly, and more troublingly, the disappearance of the future meant the deterioration of a whole mode of social imagination: the capacity to conceive of a world radically different from the one in which we currently live. It meant the acceptance of a situation in which culture would continue without really changing, and where politics was reduced to the administration of an already established (capitalist) system. In other words, we were in the "end of history" described by Francis Fukuyama. Fukuyama’s thesis was the other side of Fredric Jameson’s claim that postmodernism — characterized by its inability to find forms adequate to the present, still less to anticipate wholly new futures — was the "cultural logic of late capitalism.""
I don't know if I completely buy it or if it was Mark's (and other's) inability to cope with their ideals for the future not panning out. But it is interesting to think about, and I do see glimpses of it in most art forms. I don't know enough history to know how common this repetition of, and nostalgia for, the past is though. Perhaps a flurry of imaginings of a "new future" was the anomaly.
The replication is the new culture. Pop culture has been refrigerated and reheated for the past 2 decades , and it will be gone with the boomers. The new entertainment will be new , both in medium and in message.
Someone at work shared this and I thought it was hilarious and awesome. A number of other people found it disrespectful or gross. I'm sort of fascinated at the values divide.
I can see why SAG/ASTRA, or voice actors, aren't going to like this technology, but it doesn't feel all that different to any other industry swept away by tech, whether that's farming, horses and carriages, mining, etc. In other words, I can see the negative economic impact it will have on some people, which we may (or may not) want to protect via legislation, but I don't see how it escalates to a moral question.
"Disrespectful" to me is flipping off David Attenborough, not getting a computer to amusingly narrate mundane things in a decent facsimile of his voice.
Voice cloning seems much harder to regulate with copyright than most other issues because 1) there's a continuous spectrum of vocal similarity 2) the tools of voice cloning are available to hundreds of millions of people 3) there may be several real individuals who could claim a voice similar to David Attenborough was also similar to theirs.
Using trademarks seems like a more realistic means of regulation. Allowing David Attenborough to sue anyone that falsely and publicly credits him as a narrator seems feasible within existing law, and extending this to cover anti-avoidance tactics like "narrated in the style of David Attenborough" wouldn't be that much of an overreach.
Didn't courts already rule that AI art is not infringing on the original works that it was trained on.
I doubt that people are gonna be satisfied with the same voices, instead you will end up with David Attenborough mixed with Morgan Freeman and Patrick Steward.
I can't wait to hear the new compelling ads using this technology and I anticipate the scandal when people realize that one of the very charismatic voices belongs to an Austrian who was born in 1933 because I guarantee that his voice will end up in the training data.
Here is an idea for someone who like playing with this stuff, personalized entertainment. Pop songs sung by a voice that is a mix of Shakira, Brittany Spears and Lady Gaga, written by Chat GPT-4.
A full year ago I heard an extremely racist AI "David Attenborough" narration where he described an encounter between an Asian man and a Black man on a city street as if he were describing aggressive animals. It 100% sounded like a documentary though, a normal person would have no idea Attenborough wasn't actually narrating the world's most racist nature film. Now that is disrespectful, and already exists.
If I recall, soon after 4chan discovered this "hilarious trick", the API access for the AI narration tool was shut off. But those moats aren't going to last forever.
I believe Tom Hanks already had some one deep fake him into endorsing some medicare scam product. I'd be more worried about that than the unsavory narration which most people probably would assume was fake.
There are millions of years of evolution conditioning humans to trust things that are familiar. It's naive to think we'll adapt effectively and quickly to extreme manipulation.
You're being a little flip but adapting to one thing usually comes with compromising something else. Hanks is just an example of trusting a leader or authority. Whether you think he's a good fit for that is immaterial, everyone will be compromised.
Asking for critical thinking is all well and good but just the _knowledge_ that it's easy to generate gobs of high-fidelity forgeries, all corroborating each other as falsified evidence, could have chilling consequences by eroding faith in institutions and promoting misinformation and conspiracy theorizing.
You can adapt and still end up worse off than you were before there was anything to adapt to.
Perhaps we need to build out a way for clients to be able to trust content. Not signed with Hanks' public key, or the key of a reputable news source? The client, my tv, or whatever, should make that obvious somehow. The industry certainly spends plenty of effort developing ways for content producers to trust clients.
But I'm probably just another tech bro suggesting a technological solution to a social problem.
A great SaaS idea! It subtly injects "glory to the creator" type messages in existing science documentaries, for consumption by faith-based institutions.
The generalized kernel of your comment here leaves me deeply unsettled.
Imagine the following:
* AI discovers community's non-trivial shibboleths
* AI detects pauses in video stream where 1) the speaker is off camera and 2) there's space for a non-trivial shibboleth to be inserted
* Convert nearly all media from what it is into a lowkey propaganda piece where arbitrary community believes that their favorite celebrities are actually closet members of their community.
Example:
David Attenborough narrates a scene of ants using pheromones to encourage the colony that they need to dig out more tunnels. Then in a brief pause adds: "They must construct additional pylons."
Most people just glaze over the line, meanwhile starcraft players are having a different reaction.
"> Look forward to: David Attenborough narrates porn.*
Note that this basic idea was previously realized as the 1999 movie "The Mating Habits of the Earthbound Human" starring Carmen Electra and narrated by David Hyde Pierce.
I think, as often, the truth lies down the middle. SAG/AFTRA and the WGA were right in fighting to not give studios free rein to use AI to own their likeness or replace them in the process. Likewise, people who say using AI tech for lighthearted fun like this is “disrespectful” are just having a kneejerk reaction at this point. I struggle to see a how this is different from say, a song parody.
Your entire friends group is sharing a really funny video with your voice reciting something, some manager get's upset about it and you have to explain to management how you had nothing to do with it.
Even if you are fine with it, not everyone will be.
How is it different than someone making a funny video where they impersonate David Attenborough? Those aren’t generally considered disrespectful, right?
Like, I get the potential scale is different. But this is some joke video, not someone trying to actually replace David Attenborough or something.
I agree and the difference is commercial rights and intent. Parody, jokes, etc are fine. Using this to replace him on a documentary, not okay. Fair use already covers this all they have to do is extend copyright protections to peoples likeness and voices
> Using this to replace him on a documentary, not okay.
I think it's legal if you're the copyright holder, e.g. Disney putting dead actors in their movies did not need anybody's permission, nor does any derivative work such as videogames as they own the image right to the characters they play.
It would be interesting if they did a Planet Earth IV with Attenbrough's voice without his consent as David plays himself really...Not a fictional character.
Right now this requires API tokens and being dependent on third party companies that will cut off your access if they decide they don’t like you.
The moment these models can run locally on the kind of cheap hardware that phone scam operations have will be the real Pandora’s box moment. (I give it 3-5 years or so)
The recently released XTTS-v2 model[0] from coqui.ai is coming very close to what ElevenLabs[1] can do. It runs reasonably fast on a recent GPU, and should also work on CPU. Requires a 3 second (!) clip of the voice you want to clone. License does not allow commercial use.
> Requires a 3 second (!) clip of the voice you want to clone.
Sure, if you want a guaranteed uncanny valley experience.
There is no way a few seconds are enough to cover all the ways a specific person pronounces things. A person's voice is much more than just the pitch and with a 3 second sample anyone who knows them will be able to tell something's off within 3 seconds.
I know the parent comment said 3 seconds, but for what it's worth, the actual huggingface page says "a 6 second clip" which, admittedly, is still fairly hard to believe but I guess twice as believable as a 3 second clip.
>anyone who knows them will be able to tell something's off within 3 seconds.
Unfortunately the targets of these scammers is usually senile old people. I'm incredibly worried about my parents, especially my bleeding heart mother.
No I suspect 3 seconds is often enough. It's not learning how you pronounce each word, it's seeing some pronunciations you use and from that guessing how you would pronounce other things based on how those pronunciations are clustered in its training data. In other words if in that 3 seconds it hears you say y'all it has a pretty good shot at inferring how you say a lot of other things.
People speak differently in different situations, based on their mood and who they are talking to. They may change their voice when quoting another person, they may alter their accent if talking to a stranger, and there's many other details you don't even think about because you're used to them but it'll immediately sound wrong if it's not replicated correctly.
Apple's announced voice imitation feature requires a 15 minute sample if I remember correctly, and you can bet it would be shorter if they were satisfied with the results.
> if in that 3 seconds it hears you say y'all
I can guarantee that no 3 seconds of text of your choosing will ever be enough to reliably differentiate between all existing dialects and accents, let alone differences between individual people.
There's a reason all the decent fake voice clips on the web are based on public figures with hours of training material - and even with those you can tell within seconds that it's fake, even if you don't know how you noticed it.
> even with those you can tell within seconds that it's fake, even if you don't know how you noticed it.
Salesmen and scammers have many techniques to get people to act against their better judgement. Urgency is one: imagine a late night voice call seemingly from a loved one saying "I'm stranded, my battery is about to die. Please send the money now -" accompanied by loud environmental noises to mask any artifacts
The accuracy or realism of the emulated voice probably depends on how "rich" the recorded material is within those 3 seconds? I mean, if it includes a single long syllable, or some phrase that contains a variety of vowels and consonants.
Could work for spear-phishing, or impersonating a widely-known trusted figure. I can't really see it working for cold-calls that pretend to be someone the victim knows (like the terrifying ransom calls), since the operations work at a huge scale expecting most people to not even pick up a "scam likely" call. Even if model tuning is free and instant, just having to find a voice clip of the person prior to each unanswered automated call would tank the quantity they're able to make.
Though, for the same reason websites always attribute data breaches to a "highly sophisticated targetted attack", I imagine there will be some unevidenced claims that this is what scammers did to them - people don't want to have been fooled by something simple.
I just can't get excited over most trending submissions here on HN for the exact same reason over the last 2 years. This current advancement of AI doesn't feel like such a "new frontier" as other advancements in tech at their early adopter phase. The internet (the closest tech invention of similar magnitude imo) had an atleast decade long "wild west" before all the big players we know entrenched themselves with monopolies and legislation.
With AI we barely begun and yet the cards are already dealt.
Even though it is going to be extremely painful, one thing that excites me most about AI is the distinct possibility that it obliterates the concept of artificial scarcity that so much of the world relies on to create money these days. IMO, AI is exposing the frail and unnatural aspect of this legally-protected rent seeking.
No, I don't think the legal system will be able to keep up, and I also don't think the infoligarchs in Silicon Valley will be able to keep a lid on the box.
Yes, I'm frightened that my skills in slinging code will be available via a simple query to a model(at least as frightened as a SAG), but I'm also heartened that if these glass ceilings are shattered that we may finally make the evolutional step necessary to break so many of us free of this amoral and inhuman treadmill of misery that simply has no reason to exist with the resources at our disposal.
Using the phrase "artificial scarcity" to describe someone's skills is ridiculous. A voice actor or artist being paid for their skills is not rent seeking.
Sure, their creation is held behind legal walls in order to create, by definition, an artifical scarcity from which it derives its monetary value. If it was not protected by such a wall, and was not scarce, it would have no monetary value.
No, I don't think the inherent value in their talent is artificial, but I _do_ think that the monetary system we've built around such things as a necessity to survive in our modern world is abberant and bound to fail.
ETA: Just to give some flavor. Do you realize how much human energy is expended in finding ways to build these walls of artificial scarcity? It's frankly sad.
I was watching a showmatch that a streamer I enjoy watching was invovolved in (ShahZaM a Valorant pro). For nearly an hour in the leadup to the showmatch we got treated to a realtime negotiation between him an the player organizing his team. From comments like "no I can't use that graphic because I'm sponsored by another PC builder", to "every dollar I get I'll increase the size by 1px", and the piece de resistance "Whaaaat?! we have casters?? I'm not even getting paid for this!", followed by a streamer-initiated 10(!) count roll of twitch ads.
When the match was over, Tarik, the guy who organized the allstar team, was supposed to do an interview with the casters but had accidentally kicked off his own 10-pack of ads and they had to embarrasingly stall until his viewers were done watching them so they could see the interview live.
It was a repugnant view behind the curtains of the rediculuous amount of work a lot of people go to in order to profit behind the artificial scarcity put in front of their skills. This is usally hidden from the viewer behind production, but happened to leak in this instance due to the slapdash nature of the event.
>Sure, their creation is held behind legal walls in order to create, by definition, an artifical scarcity from which it derives its monetary value. If it was not protected by such a wall, and was not scarce, it would have no monetary value.
That's... not artificial scarcity. By definition (and using your own words) it is actual scarcity.
A person's performance -- which is more than just their voice reading lines -- is not held back behind some imaginary wall. It's literally limited by the hours in the day, the person's ability to get from place A to place B, etc.; it's not like David Attenborough is a Santa Claus figure and can be in California narrating a documentary and in New York providing some random dude a personal voiceover for his life simultaneously.
If a textbook isn't available because it's still being written by the author, then true, that's the real scarcity of the author's time. But if a textbook is finished but isn't available because effort is put into preventing the information from being freely shared, then that's artificial scarcity.
Restricting access subject to payment is currently the most common method for extracting profit to incentivize creation, but it's not the only method and for digital goods it's incredibly inefficient.
None of that is relevant to this discussion. We are talking about someone who spent time growing a skill and is charging people who want him to use that skill for whatever they are hiring them for.
nurple said: "their creation is held behind legal walls in order to create, by definition, an artifical scarcity from which it derives its monetary value"
In my example "their creation" is a textbook, and "legal walls" would be copyright/DMCA.
There is also Getty/Universal Music/etc. attempting to enforce artificial scarcity in an area where real scarcity existed (in addition to plenty of artificial scarcity) but is beginning to be overcome.
> I don't see copyright or a record label as artificial scarcity
It's allocating resources towards preventing distribution. Maybe you'd agree that it's "direct" artificial scarcity, but believe without that necessary artificial scarcity we'd soon become worse off overall?
> I think the artist should have every right to profit from their work.
Sure - but you see that is not inherently the same thing as the specific method of profit extraction through requiring payment per access to the book/image/music/..., right? An artist receiving donations or public funding for releasing their work to everyone under a permissive license is still profiting from their work, for example. I'm not necessarily saying you have to agree that alternative methods are currently feasible in all or most scenarios.
Sure, I'd agree selling access is not the only method but given that there is a story on the front page of HN about once a week about why open-source maintainers should be paid and how the current model isn't sustainable, in my opinion shows, that relying on donations or public funding is a fool's errand and not a viable alternative.
> given that there is a story on the front page of HN about once a week about why open-source maintainers should be paid and how the current model isn't sustainable
Opinion articles on HN notwithstanding, do you not think that free and open-source software has, as a whole, been fairly successful?
Moreover, IP law doesn't just apply to art or code - evergreening of medical patents is enforcing artificial scarcity on information that could save lives. Textbooks that could teach millions being restricted to the few thousand that pay is also going to have a huge hidden impact.
I don't know if there are viable alternatives for all scenarios, but given how much is on the table if artificial scarcity can be replaced - I think it'd be wrong not to actively try, encouraging any potential alternatives, rather than just being resigned to the idea that this much waste is necessary. At the very least we should oppose the attempts to enforce artificial scarcity in areas that currently managed to do without it.
>copyright or a record label as artificial scarcity.
Lets say I have a plant and I grow that plant with a bunch of other plants and make special seeds. This isn't a common skill set. Now lets say I sell you those seeds and you grow plants with them.
Now, which seems more natural.
You take the seeds of those plants and grow more seeds for yourself.
or
I prevent you with laws on taking those seeds and growing more plants.
When you put something out into the world, it is completely natural for people to take those objects and do what they want with them. It's only with more modern humanity that we reserve rights for the person selling after the sale.
No given individual has 'every right' to anything.
Virtually everything we humans do was inaugurated by those who came before, and our wholesale, liberal replication of those prior actions defines our very existence.
I think what OP is saying is that once AI can replicate these performances (and my coding skills) these things will no longer be scarce. So maybe it's not artificial yet, but it will soon be. Assuming the AI itself doesn't get exclusively owned by companies that implement their own artificial scarcity walls.
Allow me to share my own recent predicament, on which I'm still deeply cogitating, and from where a lot of my own frustration flows.
I am a content creator, I write software and build compute systems to run it upon. I was RIF'd last week and, if I'm really conservative, I have a year of expenses in savings.
I very much believe in OSS exactly because it allows me, not to tap into a monetary stream, but to overall better our human community by sharing my work. IMO, this is a core and fundamental belief of OSS and it frustrates me to no end to see articles like "OSS is going to die without funding", the two concepts are diametrically opposed.
So many, yourself included, have this perverted view of the world where there's no way it could exsist without our current monetary system, no alternative that would be more efficient, more fair, more equitable. I just refuse to believe that's the case.
I do not bemoan other content creators their right to generate monetary revenue from their creations, in fact they have no choice. But that doesn't preclude my opinion that it should change.
Indeed, what do I do now? I have 1 year to create a revenue stream off of which my family and I can live, or I need to go back to an industry that has corroded my soul in ways that sometimes make me wonder if it's even worth it anymore.
The most effective path for me to generate sustained revenue is to abandon my desire to improve humanity by sharing and turn selfishly inward, all or in part, to create a Rube-Goldbergian golem of artificial scarcity around my "content". I have to build a payment system, I have to protect the code at rest and at runtime, I have to build licensing and consider segmentation, I have to market, I have to build systems around revoking access to those who lose their own monetary stream and deal with their data, etc, etc, etc.
I know what it takes to build whole companies around systems like this, I've done it from the ground up on more than one ocassion to great success (at least for those who held the keys to the artifical scarcity doors that they paid me to build in front of my own work), and if there's anything that feels like more of a waste of my life than playing technobabel whisperer to clueless middle and upper-management, is considering building something like that again, even if I'm the ultimate keyholder this time.
In a world where we do not have to fight like dogs for table scraps, all of this wasted time and effort falls away, and instead all of us content creators who skulk through it for the love of craft, can instead focus solely on the betterment of our community as a whole. We live in a world with the resources to realize a reality like this, but IMO we have all been blinded by a masterful twisting of our inbuilt self-preservation instinct by the ultimate farce in artifical scarcity: money.
> So many, yourself included, have this perverted view of the world where there's no way it could exsist without our current monetary system
I don't appreciate this assumption; lumping me into a group of characters based on nothing I've said / something you have assumed, based on no evidence, is quite counterproductive to a healthy conversation. If you're going to try to make an argument at least base it on the content of my post and not some inference you've made based on nothing I've provided.
I'm sympathetic to your situation, and do wish you the best in staying on your feet. And I am quite, quite far from a capitalist / capitalist-apologist / what-have-you.
It sounds like you're talking toward using AI tools to effectively "duplicate" yourself, to reduce the scarcity of your own time. I hope I'm interpreting that correctly. I'm 100% on-board with letting the computers do the work so we can all go off and enjoy our lives. It's dismaying that "productivity increase" means "do more in the same amount of time" instead of "do the same in less time and then go live."
This is still a far cry from duplicating the creative work of others (cloning a performer's ability, potentially for profit), and it's still got nothing to do with actual "artificial scarcity." Again, this is real scarcity. There are only so many hours in a day. You only have so much brain power to expend within those hours. These aren't pro-capitalist statements and I'm not insinuating that anything follows from them. I'm only being pedantic here. :)
So now we’re just applying incel logic to every aspect of human creativity and expression? Do you also think that people refusing to make adult content of themselves for you is “legally protected rent seeking”?
I don’t think it’s all rude or off-topic. It’s just as easy to make deep fake pornography as it is to clone voices, so if someone claims they’re being unjustly prevented from having the one one under this insane logic, why not the other?
Absolutely right. Copyright can be seen as a kind of nano-monopoly and is not truly capitalist. The rise in prices due to stagnation does not foster innovation. While capitalism has its merits, it does bring hardship to those less fortunate. It is essential to eradicate all forms of scarcity, thereby eliminating the necessity for an economy and finance, and subsequently, the need for politics.
Taking this a step further, and being facetious, I'm thinking of starting my own network! It's all me, all the time. Sometimes I'll be a nature show, sometimes I'll be in a spy movie, sometimes I'll have the guys from Mystery Science Theater 3000 cracking jokes at how dumb I am. Can't be that hard to do what this guy did, only in twenty-four different genres, one for each hour of the day. Turn it on, let it play.
I think it would be fun. Having said that, can you imagine somebody with mental illness, perhaps that makes them paranoid, having an AI interweave them into a spy thriller or murder mystery?
People already took computers too seriously before ChatGPT and the like. Having some official-sounding narrator, somebody you know from previous appearances, helping you out in your fantasy life?
As much fun as this is going to be, it also bothers me quite a bit.
The author just casually glosses over this point and never returns to it
> One of the reasons I find these programs scary is that they seem to want to murder humans. They talk about it a lot, even when you ask them to be nice.
I was not allowed to make a meme of me with a gas can in the datacenter lighting the servers on fire as it was decommission time. It instead suggested i give it a goodbye party. Funny and a little scary at the same time.
No, I used GPT-3 before ChatGPT was even released, and it was decent at making jokes. It even wrote some dialogue between various artists and poets ruthlessly insulting each other, but unfortunately I forgot to save it. ChatGPT is very similar, but with a chat interface and much more aligned, so I don't see why 3.5, much less 4, couldn't write decent jokes.
One area where this kind of technique would be a game changer, I imagine, is for blind people to have their surroundings narrated in real time. For example, through a pair of glasses with a camera for input, and an earphone to listen to the continuous narration.
I heard about an online service where people can sign up to help the blind by describing visual scenes that they capture on their mobile. There must be companies out there developing a similar service entirely run on machine learning.
> I heard about an online service where real people can sign up to help blind people by describing visual scenes that they capture on their mobile
For anyone curious, it's called "Be My Eyes". When there's a request, it will send a notification to a batch of people. I get about one notification per month, often I'm unable to answer it right then, and when I do I occasionally get a notice that someone else answered first. Tasks I've helped people with were things like going through an unsorted stack of bills and telling them the denominations, holding up a few T shirts and asking me which is the blue one, reading a pressure valve.
Would recommend signing up. It feels good and there's no pressure to answer since it sends out the request to many people at once.
I know yours is a small comment in a huge sea, but how would the comments know what the blind wanted to see?
If I could describe what i can see now it would be a long list left to right. Do i include the sunlight reflecting off an industrial chimney? Is the chimney important?
I do enjoy hallucinogenics which, to me, remove the very filters that keep us sane.
As much as I love the problem, filtration will always be the problem, unless the environment is extremely controlled (like an art gallery), but even then...how to describe the Mona Lisa. Let alone its frame.
That does sound like an interesting and tough problem, how an intelligent agent can know what is relevant to the listener, and what to filter out.
I imagine it would work similarly to how human brains do, or at least how we think they do. To notice the "big picture" first, recognize large shapes then smaller details, or faster moving things then static things.
> remove the very filters that keep us sane
Haha, yes, I know what you mean. If our brains paid attention to everything coming in through the doors of perception, our mind would be flooded with information, way too much for practical purposes. Occasionally it is helpful to access such a mode of experience, but the filters are necessary for daily life in society. We can't be going around in rapturous ecstasy of gnosis all the time.
A related problem might be too much filtering, or the wrong criteria for the filters, missing what's important.
I imagine this could be solved by a conversational interface, where the listener talks with the ambient narrator to zoom in/out and focus on details as desired.
It’s a hard problem for sure. I guess You fine tune to their preferences. Everyone gets the “Giant bus headed straight towards you!!!!” Alert. But let’s say you’re at the grocery store, maybe it’s clever enough to read the aisle descriptions for example.
Quoting another user from an adjacent topic[0]; I for myself believe in 1., it seems generative AI tools is slowly becoming popular as lesser substitute to human arts, wherever there has to be data but any noise will do. No one listened to low-effort Text-to-Speech content pre-AI, some did for high-effort TTS content, and in neither case it was because the quality of synthesis was bad or good.
> I think Max Tegmark said there's 3 main camps of people talking about AI:
> 1. The people who think AI won't really work (at least in some fundamental way - it might make some stochastic melody but it won't have any creative spark to it).
> 2. The people who think AI will be able to do everything better than humans (in every possible way, at least as far as a mere human can tell) and that this will be great.
> 3. The people who think AI will be able to do everything better than humans (in every possible way, at least as far as a mere human can tell) and that this will be terrible.
> existential question: what does it mean to be a human?
One perspective that might be helpful is that humans have for many millennia pondered this question, so it isn't new.
We have a more exacting (but not perfect) definition from a biological perspective. We also have philosophical theories about personhood[1] and mind[2]. One can hypothesized that these, and other characteristics, such as the emergence of culture, etc. also exist in other entities (whether on earth, alien, or technology like AI), so perhaps humans can be though of us as occupying a region across these different dimensions that represent a unique combination. But we don't have exclusivity over a particular dimension (like reasoning, for example).
Another perspective that might be helpful is that even if there are other entities that are very much like humans, it does not take away YOUR and MY experience of being human or the value and joy we derive from it. If anything, it can help us understand ourselves better. This introspection ability is another one of those characteristics that humans have.
I agree but so far mainly philosophers & intellectuals used to ponder. With advancement in tech many believed that intellectual or creative work is safe. I think now this is a jolt to a lot of white-collar folks who are looped in with the recent advancements in AI.
Its not that intellectual or creative work will vanish but a lot of the current work will be commoditized.
Voices are not copyrighted and do not require "authorization". If my nephew wants to go around sounding like Goofy all the time (as he does) he is free to do so, though it may cost him invitations to dinner - it'd be the same if he was using Attenborough's voice.
The counter argument is that the basic idea behind copyright is that the expression of personality through the way you formulate ideas is a part of “you“.
That somebody would start talking like you, or looking like you, and enrich himself by essentially cloning you just has historically, never been a problem serious enough to create legislation. Just as there was probably no copyright law before Gutenberg.
Little did he know, his life would be as mundane as the one before.
(Edit: I wrote this but as soon as I did it sounds rude. Didn't intend it that way)
I haven't seen any sort of argument for why it would be secure. I guess it's probably no worse than the other common banking practice, one-factor auth using a 4-6 digit numeric only password.
Loved it, I thought that was quite hilarious! I can see how people get a little upset about it though, but in my mind this kind of content is just unavoidable.
I myself have been playing a lot with this kind of tech lately by creating realistic clones (voice + videos) of YouTubers and tiktokers. I think it has potential to go viral quite easily. I never published any of those though, because of ethical concerns. Just a question of time before other creators do this at scale though.
I almost did that with cloning real TV newscasters from bbc / fox / cnn, etc... but decided not to. But I did create a 100% "fake" journalist for an app that lets you create breaking news on the fly [1]. The result is still a little rough, obviously, but I can already tell how in just a few months it'll be impossible to distinguish a clone from a real person.
You're correct that this "creative re-application of cultural standards", along the lines of Dante / Virgil, is what's going on here.
There are some key differences. For one thing, Virgil and everyone who knew him personally had been dead for 1500 years, while David Attenborough is still alive, and those who know him personally will live for another 50 years. For another, it was absolutely clear to everyone that Virgil didn't actually say the things Dante put in his mouth; whereas as deepfakes improve, it will be more likely that people think David Attenborough actually did say whatever has been put into his mouth with a deepfake.
The whole thing's a mess. It'd be hard to draw a bright line between the crazy bullshit medievals* literally believed about Virgil, and the poetic, metaphorical bullshit Dante wrote about him.
(*I actually had a Latin professor in the 21st century sincerely advocate that Eclogues interpretation. There's a mode of thinking behind this stuff that, to put it tactfully, confounds people about what's the difference between fiction and reality).
"unauthorized" is a weasel word because it implies that it should be necessary for the developer to ask David Attenborough for authorization to create a clone of his voice using AI.
I wouldn't call it a weasel word because it describes accurately the fact nobody gave them the authorization to use the likeness of Attenborough, and there are laws to protect people's likeness in multiple jurisdictions (https://en.wikipedia.org/wiki/Personality_rights) that might require such authorization (whether they could be applied in the case of AI remains to be tested, but apparently most of them do not question the "means" to impersonate but the act of it).
It raises an interesting discussion about this though, how is this different from someone making an impression of him and doing a comedy show of it for example. They also had to listen to him for hours to reproduce the voice correctly, are they now infringing some copyright when they use this voice?
To me, as long as you clearly state that this is an impersonation/parody, and there is no deception, I do not really see a reason to protect this, even for music. But I'd like to hear arguments for a protection and how it would be enforced, how do you judge of the likeness, how do you know/prove which source material was used for training for example.
How is it different than an impersonator? 1. Let's say there's 1000 people that could, or would, want to truly impersonate him in the world as a job. Now there's 8 billion who don't need rest, who don't need to practice, who won't be indistinguishable from the real thing.
Depends on how it’s represented. If you booked David Attenborough and an impersonator showed instead you’d be annoyed. If TV shows started using an impersonator instead of Attenborough but said it was him both the audience and Attenborough would be annoyed when found out.
I'm not sure what you mean here but let me make my point to the "no big deal" proponents.
Would you be ok if AI was used to impersonate you across the internet? Would you mind if I made a twitter account in your name and phone calls with your voice or upload to Youtube videos of yourself? I wouldn't do anything illegal or to make any money. It would just be entertainment impersonating your likeness and saying whatever I would want you to say.
It is ambiguous because it makes it sound like we are talking legality here, but there's nothing illegal about making a project like this if you do not profit from it. Therefore, it is a weasel word.
If this could be created without using recordings of his voice to train the model, then sure. I agree a "sound-alike" like that is fair game.
But since actual recordings of his voice were used in order to train the model to recreate the sound, inflection, cadence, etc...it should be required to get authorization to use those recordings for this purpose.
I mean, yeah? That's also something we need to seriously consider. These tools have the ability to copy the labor and property of others in a way that is economically important and socially challenging, and we don't yet have a legal framework to handle it.
One of the popular sentiments here seems to be that training AI should not be treated any different than a human understanding & learning the source material.
Going by that logic, can the AI here be considered a "person" performing mimicry after being "inspired" by Attenborough ?
Should an actor playing a public figure in a biopic have to get permission to study the subject's voice from copyrighted sources in order to reproduce it for the film (without AI)?
Can actors audio-visually mimic their characters with increasingly convincing realism, and can they be artificially scaled, duplicated, extended, and applied for just the cost of a bit of hardware and electricity?
It is not sensible to hand wave the "human like" learning process and ignore economic reality. This area requires careful thought about how our notions of individual rights apply to new technologies, and what kind of economic system can exist as a result.
Does fair use apply here? (I don't know) It's not like the author narrates a full-blown wildlife movie with the voice of David Attenborough. It's just a funny demo which doesn't depend on the specific voice (could be anyone), and also makes it clear the voice is synthetic.
If you use a small clip of Brad Pitt in a 10 sec ad without his permission, he will probably sue you and win. So I don't think fair use is going to protect you.
Hold on, are we looking at the same video and Twitter message? [1] He clearly states what he's doing, and in the video he launches some program that appears to speak in Attenborough's voice. How clear is clear enough?
By this metric I should not ask for permission to train a model on your liking and share with your wife generated pictures of you having sex with other women
(Dune series spoiler)
For example, Duncan Idaho keeps being brought back with cloning tech, over and over, for centuries, because Leto II just likes him. Over and over, the same man keeps being put by the (nearly) immortal emperor's side...
I ask: are we stuck as a monoculture, not making new, interesting things, but simply replicating older ones?
The technology is novel and interesting, but culturally, aside from the concerns about annoying the people being replicated, it seems a bit stuck.