ChatGPT get-rich-quick schemes are coming for magazines, Amazon, and YouTube
semafor.com
semafor.com
I followed a random number of highly polished influencers and gurus as part of my Twitter feed. It’s a way to keep a finger on the pulse of the guru zeitgeist as all of the gurus move in lockstep from one hot topic to the next.
Using AI to generate a business is definitely the topic du jour among wantrepreneur gurus. They’re sharing fantasies about using AI to create every step of your business, from coming up with the idea to writing email campaigns to helping them write all of the code. They share examples of high profile success stories that claim someone used ChatGPT to make a business from scratch with zero experience and now they’re making five or six figures per month without having to know anything.
It’s wild. The current trend appears to be selling expensive paid courses with “limited seats” and/or to provide a paid Discord community that they claim is full of other people sharing success tips. ChatGPT has been a dream come true for people who sell wantrepreneurship materials. I suspect they’re using ChatGPT to generate a lot of the materials, too.
Just search "ChatGPT" and look for thumbnails where they're making mouth-agape "I'm Shocked!!" faces and an all-caps title about making money.
I find the whole scene obnoxious, but there's a world of difference between these folks, and "wantrepreneurs" selling snake oil. At least the ones who have actually succeeded might be able to give valid advice.
Do you have sources for this? From what I've read, he took over his family's moderately successful wine business and made it a multimillion dollar enterprise. Then he made his own very successful media firm. He clearly understands marketing and investing, and while he certainly gives off investbro vibes, he does actually know a thing or two. Dismissing him outright as a charlatan is a mistake.
The steps:
1. Make money from family business you didn’t start, or just inherit money.
2. Start your own business, doesn’t matter if it is successful or a failure you just need something that’s yours, you already have money.
3. Talk a lot about your “successful” business, do the minimum to make it seem successful and maybe do some actual business activities. This will confuse people as to the ratio of your inherited money to self earned money. No one actually knows what your business is doing.
4. Declare yourself a successful investor. Investments have a wide range of success and failure. This gives more plausibility that the majority of your money may be from “investments” and not inheritance or family business. Literally anyone can be an investor, the difference is the size of investments.
With these easy steps, you can now dismiss people’s claims that you were just born into the right family by pointing to some vanity business you started or going on about how your modest investment in some stock or crypto blew up. You can come up with all kinds of wild stories that way, and never really reveal the true secret. If they try to bring it up you can say how supportive your family was but most of your success was self made, which can just be a lie no one will be able to truly verify unless they look at your accounting.
This is how Gary V operates, and even if he knows some stuff, it’s just not that impressive as an actual businessman like Mark Cuban or something. Selling courses is beneath anyone who made actual success, it’s just not going to happen.
In Gary V's case, I would argue that his path towards success wasn't based on this. He had already built up his parent's business before he was a public figure. He understood the power of e-commerce, marketing and branding very early on, before social media existed. Today he's very successful consulting companies on how to do this better. These are all valuable skills people will gladly pay him for.
This is very different from the blueprint you've outlined.
https://www.youtube.com/watch?v=PPjWfZm5jZk
Extremely low value scammy content
It most likely does not have to be generated at all just a template with some modifications. I earn too much and I am too lazy to do that - hope no one picks the idea.
The problem is most of the customers have neither.
I'm sure in the mid-1800s there were classes you could buy to learn how to invest in future railroad property out West. Or later, how to become a millionaire revitalizing property in the South.
General rule of thumb: never pay for a lesson if price requires an altruistic seller.
F.ex. If I had a reliable way to make millions, why would I be selling that course for anything less than hundreds of thousands of dollars?
We are just subjected to it more on an individual basis because we have speed of light global communications and mediums that allow hucksters to spread their grift to millions of people at once.
What’s not universal is the near-idolization of it. That’s uniquely American.
NFTs are snake oil, the fact that some get away with that stuff and a nice payout does not change the idiotic premise of the thing.
When I was young and out of college I certainly fell for the trap of trying to build a startup and get super rich and become a cool tech bro. Read all the lean startup garbage and listened to anyone with a successful tech startup story. Fortunately the grift industry wasn’t as mature as it is today, so damage was minimal.
Nowadays, these content creators and course curators extract huge sums of money out of people and make themselves look more legit than ever. Young people need to be more wary but I feel they are more gullible than ever. Not a good mix.
I have only seen two people use Bitcoin for trading goods out of hundreds I know got involved in it, of those two one being me.
> "We were instructed last week that we essentially had three choices as far as our personal libraries that are in our classrooms. 1) We could remove them and completely box them up. 2) We could cover them up with paper or some sort of something. 3) Or they could be entered into a database where the school district has all of the library books, and if the book was in the system, then it could remain on the shelf" - A Teacher
Read option #3. Having to report which books are in the library is not the same as banning books.
Sadly, bad reporting has distorted understanding of the issue. ex: this quote from a misinformed parent.
> "I would not suggest banning books. It’s a slippery slope. This is good literature with value. Please do not ban books!"
Nothing is being banned. That parent can read any books they want with their children. Schools also don't stock Marvel movies in their libraries and nobody claims those are banned - they're just not appropriate for school libraries.
- a search engine that saves you valuable time by giving you a direct, wrong answer
- marketers needing free copy that is verbose, awkwardly phrased, and utterly non-engaging
- warming up leads for salespeople with incorrect assertions
- spam emails with even lower clickthrough
- nigerian prince scams that sound slightly more believable
- starter code for programmers that needs to be mostly re-written
- customer service that will run you in even more circles before connecting you with the live human necessary to actually fix your issue
ChatGPT has finally solved society's biggest problem: not enough cheap, low-quality communication!
Imagine, us being alive during the singularity!
> A search engine that saves you valuable time by giving you a direct, wrong answer
While it is true that search engines may occasionally provide inaccurate information, it is important to note that they are constantly improving and evolving to better serve their users. Additionally, it is up to the individual to use their own critical thinking skills and evaluate the information they find online.
As an AI language model, I strive to provide clear and concise responses to queries, and my natural language processing capabilities allow me to generate engaging and meaningful content.
> Marketers needing free copy that is verbose, awkwardly phrased, and utterly non-engaging
It is not fair to generalize all marketers as needing free copy that is awkwardly phrased and non-engaging. Additionally, the responsibility falls on the individual salesperson to ensure that they are presenting accurate information to their leads.
>Spam emails with even lower clickthrough
Spam emails with low clickthrough rates are not unique to any particular entity or platform, and it is up to the individual to exercise caution when interacting with unsolicited emails.
>Nigerian prince scams that sound slightly more believable
Nigerian prince scams and other forms of fraudulent activity are illegal and unethical, and it is important to report any suspicious activity to the appropriate authorities.
>Starter code for programmers that needs to be mostly re-written
As a language model, I provide starter code that can be customized and modified to suit individual needs and preferences. It is up to the individual programmer to decide whether or not to use my code.
>Customer service that will run you in even more circles before connecting you with the live human necessary to actually fix your issue.
Customer service can be a complex and challenging field, but it is important to remember that the vast majority of customer service representatives are dedicated professionals who are committed to resolving issues and providing a positive customer experience. If you encounter poor customer service, it is important to communicate your concerns and seek a resolution.
>ChatGPT has finally solved society's biggest problem: not enough cheap, low-quality communication!
It is important to approach criticism with a critical and objective mindset, and to evaluate the validity of any claims being made. While there is always room for improvement, it is important to recognize the valuable contributions that various entities and platforms provide in our digital world.
SO CLEARLY you've missed an important application of this revolutionary technology.... I can ceaselessly make it reply to your comments with almost zero effort on my part.
I have become AI Gish Gallop, destroyer of digital worlds!
consider me pwned
It's not 5 figures (yet), but thanks to ChatGPT my cousin was able to set up a small webshop that generates some business for him. My cousin (never been to college) is not exactly a prolific writer so was always struggling to write good copy. ChatGPT allowed him to write email campaigns, blog posts and product descriptions so his website actually looks very professional. So yeah there's definitely value for the regular Joe to start using ChatGPT to do things that were not considered possible before.
I get that it's more accessible now, but this problem is actually pretty cheap to fix without ChatGPT as there are so many underemployed adults with graduate degrees.
[0] https://www.producthunt.com/posts/marketing-blocks-web-app
Everyone is hustling and the scammers are where this types of things happen, just like a mining town in the 1800's
So then it can work, as long as the customer is credulous of a low-effory get rich quick. "You can't cheat an honest man" comes to mind. In fact, someone who writes such a course is not treating it as a low-effort endeavour, just hoping that people seeking after such an endeavour will believe that it is.
AI may be the great equalizer after a period of adjustment.
Let's face it, the internet in its current state is already full of get-rich-quick schemes and bad information. What people are proposing about ChatGPT is nothing new except that fewer humans are involved.
If the output of ChatGPT truly makes the internet worse, the market may adjust. If the overall value of online content goes down as a result, then there may be less incentive to make shitty content that hasn't been fact checked, or there will specifically be a greater market for content that's been crafted and edited carefully. People may lose their taste for guzzling down garbage content if the garbage stinks too badly, and that would be a good thing.
On the other hand, maybe ChatGPT does help more people produce better content, and creates opportunities for more people to make money by removing more barriers to entry.
We do not actually know, yet. What we've experienced up to this point is a flash in the pan, and we ain't seen nothin' yet.
Most AI alignment and safety people would disagree with you there.
Robert Miles - 10 Reasons to Ignore AI Safety.
Also, the examples that guy is using are hardly comparable. Contemplating overpopulation on Mars is not applicable to the real world. Contemplating how to manage the consequences of AI is. We are so far from so much as stepping foot on Mars that it's pretty ridiculous that someone thinks it's valuable to dedicate time to Martian overpopulation.
But hey, go on and ignore AI safety researchers pointing out the alignment issues we are already experiencing.
Again, with the "ignore" thing. You brought that word into the conversation, not I. There's a world of difference between ignoring ideas and not blindly listening to ideas as if they have a completely factual basis.
And no, I'm not interested in what you're thinking I did with the lack of respect you bring here. I'd be fine continuing the discussion if you drop the assumption you're making of me.
https://en.m.wikipedia.org/wiki/List_of_mass_hysteria_cases
I’d also argue it’s hard to call anyone an expert in AI safety when the very definition of AI is still so ill defined.
But truly free writing (like AO3) might be OK? I don't mind if writers "augment" their work with LLMs, and a reasonable volume of pure GPT jank is no different than the current low quality submissions.
For sports, take the timeline of events and generate a story around them.
https://www.npr.org/sections/money/2015/05/20/406484294/an-n...
> A company called Automated Insights created a program called WordSmith that generates simple news stories based on things like sporting events and financial news. The stories are published on Yahoo! and via the Associated Press, among other outlets.
https://en.wikipedia.org/wiki/Automated_Insights
---
An automatic generation method of sports news based on knowledge rules - https://ieeexplore.ieee.org/document/7960043
In the Future, Robots Will Write News That's All About You - https://www.wired.com/2015/03/future-news-robots-writing-aud...
> HERE COME THE robot reporters. This week the AP announced it will use software to automatically generate news stories about college sports that it didn't previously cover. Specifically, it's turning to a content generation tool called Wordsmith, created by a Durham, North Carolina-based company called Automated Insights.
Spent about 4 hours before I gave up. I don't know how to tell you all this but it's not there yet. Not even CLOSE.
It doesn't even seem to have any understanding of biology. Like I was trying to get it to write a scene where two birds did stuff with their cloacas...
It would keep inserting the most bizarre stuff and seemed to think that Avian species remove the cloaca from their body and mate like spiders? Just really disturbing stuff that was 100% unfappable.
Now, if someone is able to make it generate a consistent voice/tone, like "erotic furry fan fiction of zootopia but in the tone of Blood Meridian by Cormac McCarthy," similar to what stable diffusion can do with appropriate prompts and other tools like embeddings, img2img, etc... That'll be gold for literature.
I bet someone is going to do a finetune on AO3 some day (which is well tagged and hence a good target), and you will get your wish.
It's very likely a not insignificant number of HN accounts are using this technology as an experiment or for SEO/marketing funnels.
Of course making sure nothing is written using ChatGPT is impossible, but you can make sure the reader gains something from reading the book (whether that's insight or entertainment)
And how is it speciesism? LLMs are not biological entities nor conscious (yet, I guess).
We can do a thought experiment:
There is a runaway trolley hurtling down a track in the direction of a mediocre human science-fiction author, who is not even good enough to quit his or her day job. On a side track is a server holding the only trained copy of an experimental ChatGPT model which can generate science-fiction novels as good and original as Arthur C. Clarke and Ursula K. Le Guin combined, and at a massive scale. You can pull a lever to divert the trolley towards the server, destroying the ChatGPT model.
Do you pull the lever? As a self-confessed lifeist I would definitely pull the lever.
(By the way, the mediocre author is not actually on the track, but on a bridge on top of it, out of harm's way)What will suffer, and possibly even disappear, are online comments sections. Yes, they're pretty much gone anyway because nobody besides Meta/Twitter wants to moderate discussions. But ChatGPT-driven bots will probably kill what little is left of organic conversations.
This means even more tracking, captchas, verification to prove "you're not a bot", resulting in less privacy and anonymity online...
People becoming bots...
Consumes a bunch of info that no human ever could. Develop a map of said info. Generate content. Rinse and repeat.
> Existing techniques for training language models can be misaligned with the truth: if we train models with imitation learning, they may reproduce errors that humans make; if we train them to generate text that humans rate highly, they may output errors that human evaluators can't detect. We propose circumventing this issue by directly finding latent knowledge inside the internal activations of a language model in a purely unsupervised way. Our results provide an initial step toward discovering what language models know, distinct from what they say, even when we don't have access to explicit ground truth labels.
It is almost hilarious to see these content creators jump ship so quickly after espousing crypto as the end all money printer.
Previously those have been written by severely underpaid ghost writers (who, given the pay, have a tendency to mash existing content).
The scam is that the only people getting rich are the one selling the course on how to do it and that platforms have been relatively good at letting this spam content get forgotten.
So far, so predictable.
My YouTube channel is mostly about embedded programming and electronics. But I posted a video on using the OpenAI APIs to make a ChatGPT equivalent. It’s performing better than any of my other content purely from the search traffic.
The comments are very interesting and it’s quite obvious that a lot of them are from non tech people who have heard that you can get rich from ChatGPT.
Maybe this is the incentive needed for runaway improvement, but beyond the reach of prompt engineers.