The whole article feels to me like it’s generated by GPT-3 based on a few prompts. There’s a reference to Forbes article not linked, but otherwise zero things backing up this 1% claim.
The thing that makes this very suspicious is the continuous repetition of the same content, the anonymous writer and the fact that the only other article in this publication is about a tool that’s even better than GPT-3.
GPT-3 certainly has the effect that I have a hard time trusting that anonymous articles that are repetitive are not AI generated any more…
Update: I posted the exact same comment on the article and the author deleted it within two minutes. instead of responding. So yes, it’s likely I was on the money. I re-posted the comment. If it’s not there, you know that this comment is uncomfortable for the author for some reason.
Update 2: my second comment was removed within minutes as well. There’s a commenter claiming they are the author saying they used ChatGPT to generate parts of the article. Does not explain why they keep deleting my comment and not disclosing that this article is AI-generated.
Update 3: posted a third and final comment asking the author to not delete this comment and answer if the article was verenigde by ChatGPT. Comment also deleted within minutes.
This all underscores how it’s becoming hard to trust anonymous authors even now, and how this will just get worse.
If a GPT3 article can generate so much discussion on Hacker News, without most people realizing we are arguing about the output of an AI, GPT3 is ready to go mainstream.
Thanks for the updates, can you add another update with the author's response? Whether you agree with the argument or not it's relevant they have shown face and have justified their actions.
I posted on this thread earlier too:
Hi! I am the person that wrote this, and used GPT-3 and a few other writing tools to help me wordsmith it. All the points, however, is original work and not AI generated. I am not a native English speaker, so I have been using these tools to avoid awkward sentences/paragraphs. (Clearly this has not been the outcome I was hoping for)
As for the erroneous citation to Forbes: there were two links, one to Forbes and the other to Digiday to backup the next point in the outline I wrote. While transferring the content from Huggingface to the substack editor, I missed that in the proofreading.
So why delete my comment on the publication instead of addressing it? Then why delete my second comment? And why not make it clear the article is generated by ChatGPT, at least partially?
Also, this is a good example on why anonymous accounts writing content will be trusted less. You say you used ChatGPT for parts of it. Unclear on what “parts” mean and how much input you had, versus what the AI wrote.
It’s a reason for people to stop reading anonymous authors, or articles that don’t make it clear that it’s not an AI writing part of the article.
I’ll be honest: I feel duped reading a wall of text to realise it’s at least partially generated, and this whole article could have been the prompts you used to generate it.
I am not as comfortable having a back and forth on substack than I am on HN. I am not sure as to why people who use tools like [1], [2], [3], [4] have to declare as such that the parts of the content may have been edited with help from computer tools if the end goal of conveying the information has been satisfactorily achieved. This seems like a rehashing of the "AI generated art is not real art" debate [5] again.
I mentioned, I am not as comfortable having a back and forth on substack than I am on HN. That is why I deleted the comments on substack, as they were not relevant to LLMs (or "GPT-3" as you like to call them)
Ignoring for the moment the accuracy of that tool, even if the prose is 99% AI written, that doesn't meant that the result is fake. Maybe their model wrote it.
100% had this thought because of the repetition. It generally just feels like it makes the same point three times without adding anything new in any of them.
Hi! I am the person that wrote this, and used GPT-3 and a few other writing tools to help me wordsmith it. All the points, however, is original work and not AI generated. I am not a native English speaker, so I have been using these tools to avoid awkward sentences/paragraphs. (Clearly this has not been the outcome I was hoping for)
Citing without linking is something I've seen ChatGPT do. It knows this sort of article should contain this sort of citation but doesn't know or care what to actually link.
The second half of the article seems more authentic though. I'm not sure ChatGPT would come up with "the middle class comes from economic friction" (paraphrasing).
“ The creator economy, once hailed as the solution to economic inequality and the savior of the middle class, is facing challenges that threaten its very existence.”
Not to nitpick, but did anyone really believe that becoming a tik tok influencer was a reasonable career path for more than a couple hundred people? I don’t see how most middle class people would’ve been helped by the creator economy
> but did anyone really believe that becoming a tik tok influencer was a reasonable career path for more than a couple hundred people?
Yes, there are huge swaths of the population that are, to put it politely, low information and unsophisticated. Everyone comes around to the disappointment of the situation at their own pace.
THIS. Just how far into can’t-do-arithmetic self-delusion do you need to be, to believe that “you need X thousand followers to succeed financially” jobs could actually be available to more than a few people out of every X thousand???
From a theoretical point of view, that is not a problem. Given a population of Y people, each of which are following X other people, you can get X as high as Y-1. Fan bases are not mutually exclusive, just like I can shop both at place A and B and sell stuff at my own store C.
Sure, everyone can follow everyone, 8 billion people can each have 8 billion followers, but that also means that everyone can spend about 4 milliseconds per year on each one they are following. What really matters is the time and money you are spending and that is essentially a fixed amount which you can split into a few larger or many smaller contributions. If there are many evenly spread small contributions, then no one can make a living on that, if there are a few big and concentrated contributions, then only a few can make a living.
Yeah but the size of those fan bases does have some practical upper bounds. Usually accounts follow a set of x people where x is in the hundreds whereas creators need to get to Y followers where Y is typically measured in a size orders of magnitude bigger than x. So you have a, practical, hard asymetry
It's actually a problem of time. Everyone on earth could follow everyone else. But you have 24hr in a day to fight for. You probably have about 5 hours of a heavy user's attention to fight for.
Most people are still low wage or "gig economy" service workers. Which is a big part of why these escape-the-grind fantasies are popular and have been for generations. The problem is the world we've made not how people are trying to navigate it with hope in their hearts.
I think a massive number of 15-25 year olds think it's not only a reasonable career path, but their preferred career path. They don't expect to become middle class though, they expect to become rich and famous.
What people forget is that when everyone is an influencer or a creator, nobody is.
When a youngster tells me they want to be an influencer for their profession, I simply ask them "who are you influencing, and what sort of credibility do you have in order to influence them?"....seems to get the kids every time - they haven't put 2 and 2 together. The credibility needed to influence comes before everything else, and people really don't understand that, for some strange reason.
The original influencers were people who were already doing cool stuff and just decided to film it and post it on Youtube 10+ years ago. Shit, the electronic musician, deadmau5, was streaming his production sessions online in like 2007 or 2008. To me, that's a creator - someone who creates for a profession and would be doing so regardless if social media existed or not. The guy was also an early influencer in the music tech space because he would talk on his streams "hey yeah I've been using this plugin or that piece of gear in my latest tracks, here's what it sounds like, and here's how I use it". I dig that, that's legitimate - you have a professional talking about tools used in a professional environment and how to use them.
Nowadays everyone seems to think that because a few hundred bots have followed their Instagram account, they are entitled to make millions of dollars by posting stupid dance videos or copying the latest trend.
Heh, this echos my life somewhat, I wanted to be a writer in high school and then I realized I had absolutely nothing interesting to say as I had no life experiences or expertise. So I decided to acquire both and by the time I did that I had no interest in writing.
Maybe it's just part of being a teenager, but it's amplified dramatically I think by the fact that there are countless successful teenagers / early adults on the platform and these kids have parasocial relationships with them.
There are also a massive number of 15-25 year olds that think fine art is a reasonable career path. Or sports. Or music. Or acting in Hollywood. Or writing books.
Art and entertainment has had more supply than demand for at least a century because it is a great life for self-directed creators.
I was about to say the same thing. The article's premise is based on nothing. I've never heard a serious person say or believe that content creation would be a viable and stable career for the masses.
Even the 1%'s success is shaky. It's based on fashion, hysteria, playing engagement games and algorithms snowballing them into power. It cannot and will not ever work for a huge group, that's just not how attention is distributed. It's by definition winner-takes-all.
Definitely. Back when I was young something like half my school wanted to be a professional footballer. There was a period after that where lots of people wanted to be esports gamers. Tiktok is currently _the thing_ which has taken that spot, as far as I can tell.
I think there was a trend towards "I want to be a programmer, cos they get paid loads and have free food", but don't know whether that's still with us.
I'd blame this on a mixture of unfounded optimism and, for my generation at least, literally everyone offering the career advice of "Do what you love and it'll all work out great". Nonsense but sounded plausible at the time.
It's like saying being a professional basketball player is not a viable career path because most people don't make it. The mere fact of it's exclusivity is besides the point.
Being a content creator/influencer is a viable career path. There are creators with disabilities, creators that never speak, creators that never show their face. Not everyone makes great money but for many people the money is enough.
I half agree with this article. There does seem to be a typical power law in terms of who is able to make a bunch of $ and make this a viable career.
That being said, 'winning' doesn't necessarily mean having this be your entire job or revenue source. I think we're progressively going to see more 'portfolio careers' or people holding multiple small jobs. I've seen many folks have small ish followings (say few thousand subs on YouTube), but that's enough for them to generate good leadflow for a consulting business that sustains their family. Put another way, being a creator can just be the marketing or top of funnel for other activities.
If we view it solely as a full-time job, then it probably will be a privileged few <1% that 'win'. But if it's part of a larger portfolio of work, then the creator economy opens up a lot of new possibilities for people.
Put another way, being a creator can just be the marketing or top of funnel for other activities.
Then, I would argue, you are not a creator, you are just running your own ads for your actually business on YouTube. This is not to say that it is a bad choice, you might really enjoy it, it might be more cost efficient than other ways of advertising or whatever.
As I said elsewhere, I actually hat to look up what the creator economy is supposed to be exactly. I understand it as producing online content for a living and therefore I would argue that this content should be the primary source of your income in order to make you a proper creator. Does someone in the Microsoft PR department become a creator because they are filming stuff on the campus and put it onto their YouTube channel? There is however probably a grey zone somewhere, where the content that you are producing and other stuff or services you are selling become so intermixed that it is hard to separate them.
You are welcome to use the word creator in any way you see fit, I however would not dilute the meaning in this context until it becomes a useless word. If I am forging a knife, I am creating something, if I am operating the camera in a Marvel movie I am creating something, if I am filming the wedding of friend, if I am crowing tomatoes in my garden or in a massive green house farm. For my understanding of the word it matters what you create and in which way this is linked to making a living. You are not a maker just because you make stuff and get paid for it because that applies to almost everything under the sun and makes creator synonymous with having a job. Maybe not blowing up old buildings...well, even then I am making room for new things.
Yeah that's fantastic! I'm a huge proponent of diversified streams of income, and the low barrier to entry means people can try it out with little investment and a little elbow grease. If that can bring in xx% of your full-time income I think a lot of people would call that a win.
The problem with an economy like that is it doesn’t scale up to support a modern state. Things like a universal health service (even to the extent of Medicare, Medicaid and CHIP in the US); national road, rail and telecoms infrastructure; a universal education system; social services. You need a massive surplus to pay for all that stuff, that an artisanal craft based economy just can’t support.
I share your doubt. As for sustainability, the key to that is efficient utilisation of resources. I don’t see how that can be possible without scale efficiencies.
Basically there are trade offs between population level, standard of living and sustainability. More of any one of those will require us to make do with less of one or both of the others. A more efficient economy can give us more to distribute between them.
That’s broad strokes. There are cases where some efficiencies compete with some standard of living concerns for example. It would be a poor world in which there were no artisanal crafts.
Look at wealth instead of income, and you will probably also find that "the middle class" has always been a bit of rose-coloured labelling of some slice out of a Pareto distribution. Traditionally, people have felt middle class because they were not comparing themselves to an entire distribution — but (almost by definition?) the "creator economy" has been focused* on a tail.
* Eg I can easily find numbers for mean AdSense payments, but not median, nor, more importantly median including those creators who don't meet AdSense minimums.
The idea that there is no middle class any more us utter tosh though, nowadays in the western world almost everyone is middle class, by any stable criteria.
Of course it depends how we define it. If the definition is some mid percentage by income then it would be fairly consistent over time because the absolute criteria would be constantly changing. However with those measures someone can maintain the same standard of living in society and leave or enter the middle class as demographics shift around them. If we use a more absolute measure that better tracks demographic change over time, such as having one third of income available for discretionary spending, or criteria such as home ownership or type of work done so we can better track changes over time, then more than half the worlds population is now middle class.
I actually had to look up what creator economy even means exactly and if Wikipedia [1] is accurately describing it, then it seems to essentially belong into the broader category of entertainment.
The creator economy was once hailed as the solution to economic inequality and the savior of the middle class.
How would that have worked? You can not run a world on entertainment.
Are there any entrepreneurial professions that do not have this winner take all dynamic ?
The only ones I can think of are professions with some physical constraints. Eg where the product requires the professional to be physically present. Eg surgeons, luthiers, sports therapists.
Anything that isn't scalable. Taleb writes at length about this difference in The Black Swan, categorizing professions in which there are winner-take-all effects (Extremistan) and low variance, stable, one-sample-is-representative ones (Mediocristan). As you said, surgeon (the example Taleb gives is 'dentist'), prostitute, taylor, and so on. He also notes that some professions can switch categories, e.g. musician, since the existence of recording technology the distribution became a Pareto one.
Depends. If surgeons somehow ended up working for a company/franchise then that company will hold all the power. Surgeons would have to unionize to get some power back.
I hear that something like this is happening with vets.
It's actually not Pareto's but more like "Normal Distribution" and that is something embedded in Nature/Society that one can't change or eradicate. It's just the way things work, the natural dynamics.
Yeah, is the top only fans girl really 10x or whatever more valuable than the penultimate only fans girl? Same for the arts. Same for most market segments whose products are artfully differentiated to prevent the customer from commoditising them mentally.
>Yeah, is the top only fans girl really 10x or whatever more valuable than the penultimate only fans girl?
That's the wrong interpretation. The reason why the top only fan girl gets 10x more revenue than the penultimate only fans girl is because she's marginally better, so more people flock to her. This works because her marginal cost of production is zero, so her having 10x the fans doesn't cost her 10x the work. The same applies to other professions where the marginal cost is zero. I doubt the average NFL player signing multi-million dollar contracts are hundreds/thousands of times better than the amateur player, but because sports leagues wants to hire the best, the top players get a disproportionate amount of money chasing after them.
The algorithms tend to favour entrenchment. They don't select on good or best they don't understand those qualities so they select on the basis of popular, assuming people are aware of the best. The issue with that is once a best has been selected years later the algorithm has effectively boosted them even with better options. Discovery is a real problem on most social media but especially on places like twitch.
I think it's key for creators to diversify their income and generate revenue from different streams. This is especially true in light of platforms becoming increasingly restrictive and unpredictable with their content moderation policies. Relying solely on automated ad systems as a means of generating revenue leaves the creator susceptible to the algorithm. It seems like the biggest asset that creators are creating are the communties that form around them and their content. Even for smaller creators, I've seen time and time again that all you need is one or two highly dedicated and engaged fans to make being a creator an extremely lucrative endeavour.
I've been working on a platform to help creators diversify their revenue streams and focus on developing their community as an asset on top of just creating content. The hope is to try and bring more predictability in their ability to monetize. http://sociables.com/creators
Substack is in trouble because of this. They get almost all their revenue from two handfuls of writers, any of which could afford to develop their own ‘pay for an e-mail newsletter’ script and keep more money.
Unlike some companies that look similar, I think Substack is not delivering a lot of value in terms of discovery. People come to Substack for Bari Weiss and not the other way around. It is not Medium, the low value domain of people who want to put low effort into blogging, where people are blown away because they got 70 views on a post. (Prediction: Medium’s user base will defect once somebody comes up with a platform that uses LLMs to write their blog posts for them)
Substack would like to grow to be more than e-mail newsletters but the creators and users will have a say in that.
Are we just paraphrasing a critique of the "The Long Tail: Why the Future of Business Is Selling Less of More" book [1]? Internet offered new eyes for new content but few winners remain. For sure Internet offers new opportunities to people. For example, well paid remote dev jobs outside of the center of power (SV) but they are not creators, they are workers.
The thing that makes this very suspicious is the continuous repetition of the same content, the anonymous writer and the fact that the only other article in this publication is about a tool that’s even better than GPT-3.
GPT-3 certainly has the effect that I have a hard time trusting that anonymous articles that are repetitive are not AI generated any more…
Update: I posted the exact same comment on the article and the author deleted it within two minutes. instead of responding. So yes, it’s likely I was on the money. I re-posted the comment. If it’s not there, you know that this comment is uncomfortable for the author for some reason.
Update 2: my second comment was removed within minutes as well. There’s a commenter claiming they are the author saying they used ChatGPT to generate parts of the article. Does not explain why they keep deleting my comment and not disclosing that this article is AI-generated.
Update 3: posted a third and final comment asking the author to not delete this comment and answer if the article was verenigde by ChatGPT. Comment also deleted within minutes.
This all underscores how it’s becoming hard to trust anonymous authors even now, and how this will just get worse.
If a GPT3 article can generate so much discussion on Hacker News, without most people realizing we are arguing about the output of an AI, GPT3 is ready to go mainstream.