The author links to the somewhat dystopian blog where the email sender is quite proud of their work. Their words (or perhaps that of an LLM):
> Could an AI agent craft compelling emails that would capture people's attention and drive engagement, all while maintaining a level of personalization that feels human? I decided to find out.
> The real hurdle was ensuring the emails seemed genuinely personalized and not spammy. I knew that if recipients detected even a whiff of a generic, mass-produced message, they'd tune out immediately.
> Incredibly, not a single recipient seemed to detect that the emails were AI-generated.
The technical part surprised me: they string together multiple LLMs which do all the work. It's a shame the author's passions are directed towards AI slop-email spam, all for capturing attention and driving engagement.
How much of our societal progress and collective thought and innovation has gone to capturing attention and driving up engagement, I wonder.
I remember seeing a talk from Jonathan Blow where he made a comparison: in the 1960s top engineers worked for NASA and put a man on the moon in a decade, basically doing computations by hand.
Today, we have super advanced computers and tech companies enjoy 100× times more of the top engineers than NASA ever had, and they are all working toward making you click on ads more.
Just wait. Enough of us will get pissed off that we will develop AI agents that sit between us and the internet.
A sufficiently advanced personal assistant AI would use multimodal capabilities to classify spam in all of its forms:
- Marketing emails
- YouTube sponsorship clips
- Banner ads
- Google search ads
- Actual human salespeople
- ...
It would identify and remove all instances of this from our daily lives.
Furthermore, we could probably use it to remove most of the worst parts of the internet too:
- Clickbait
- Trolling
- Rage content
I'm actually really looking forward to this. As long as we can get this agent into all of the panes of glass (Google will fight to prevent this), we will win. We just need it to sit between us and everything else.
Someone decided that marketing is now a tech problem. Artists have been replaced by software engineers. The net result is creepy AI emails.
I fell for oldschool marketing yesterday. Im moving into a new appartment in a couple months. The local ISP who runs fiber in my new building cold-called me. I agreed over the phone to setup the service. That was proper target marketing. The person who called me knew the situation and identified me as a very likely customer with a need for service (the building has a relationship with the ISP). I would never have responded to an email or any wiff of AI chatbot. They only made the sale because of expensive human effort.
And yet someone is building all those super advanced computers and AI models. Someone is launching reusable rockets into space. Someone is building mRNA vaccines and F1 cars and humanoid robots and more efficient solar panels.
The "smart people are all working in advertising" trope is idiotic. Just an excuse for people to justify their own laziness. There is an infinite number of opportunities out there to make the world better. If you are ignoring them, that's on you.
Which do you think is more important? Putting man on the moon or ecommerce? I reckon you been able to get on a device, see a biscuit ads, order one from foo.com and have it shipped to you. Think of how much tech it takes for that to happen, that is more tech than NASA built to send many to the moon, the internet, packet switching, routing, fiber optic, distributed systems, web servers, web browsers, ads, cryptography, banking online, and so on and so forth. We love to trivialize what is common, but that clicking on an ad is not an easy problem. Clicking on ads has generated enormous wealth in the world which is now bootstrapping AGI.
Clicking on ads helped with our goal to AI today. Showing you the right ad and beating those trying to game it is machine learning heavy. When was the first time we started seeing spelling correction and next word suggestions? It was in google search bar. To serve the correct ads and deal with spam? heavy NLP algorithms. If you stop and think of it, we can drop a think line from the current state of LLMs to these ads click you are talking about.
The guy writes a post about how to send spam effectively, and then offers the subscription link in the end with "Promise we won't spam you". Yes, I totally trust you...
I keep seeing these posts on HN and thinking, man, these are some smart people. Training LLMs, doing all this amazing AI stuff like this guy with the email agents and the other guy with the dropping of hats, and then I open the posts and it's just some guy making API requests to OpenAI or some similar disappointment.
When “altcoins” took off I spent a while racking my brain trying to figure out what special tech I could offer, how I could build my own blockchain, incentivize miners…
When I realized it was just dudes copy-pasting a “smart contract” and then doing super shady marketing, it was already illegal in my jurisdiction.
Nowadays, an "AI Expert" is someone who knows how to download an AI client lib and prompt the AI to perform tasks. These are people who are not even technical and have no idea how all this works, but they can at least follow a Youtube Tutorial to get a basic website working.
well, no one's going to be talking about the secrets behind LLM while the market is paying billions to own their slice of the pie.
And in reality, most software work is 1) API calls and 2) applied math. If you're not in cutting edge private tech or acedemia, your work probably falls into 1 or both categories. Modern "Software engineers" is more a matter of what scale of APIs you're wrangling, not how deep of domain knowledge you have.
This process should not require a human in the loop.
Consider:
* spammers have access to large amounts of compute via their botnets
* the effectiveness of any particular spam message can easily be measured - it is simply the quantity of funds arriving at the cryptocurrency wallet tied to that message within some time window
So, just complete the cycle: LLM to generate prompts, another to generate messages, send out a large batch, wait some hours, kick off a round of training based on the feedback signals received; rinse, lather, repeat, entirely unattended.
This is how we /really/ get AI foom: not in a FAANG lab but in a spammer's basement.
At least one sci-fi novel iirc had an AI spam filter achieve sentience, because the task basically amounted to contrastive-learning a model of humanity.
This is sort of why I feel somewhat pessimistic about AI - the inevitable most profitable usecases being so bad in aggregate for a society with almost no bounds or values other than profit. It will never be easier to waste peoples attention.
I think it depends on the context. I think there's artistic cases for it, for example I've played around with using AI tools to extract speech from its background music for use in further (non AI-based) music which I don't think is an unethical thing to do.
It's already like this for creative communities in things like illustration and writing. You will (rightly) get ostracized and blocked by your peers for using AI. It's a signal for poor quality for most people in those spaces.
Definitely interesting to see the different culture in tech and programming since programmers are so used to sharing code with things like open source. I think programmers should be more skeptical about this bullshit, but one could make the argument that having a more flexible view of intellectual property is more computer native since computers are really just copying machines. Imo, we need to have a conversation about skills development because while art and writing accept that doing the work is how you get better, duplicating knowledge by hand in programming can be seen as a waste of time. We should really push back on that attitude though or we'll end up with a glut of people in the industry who don't understand what's under all the abstractions.
News at 11, spammers use sophisticated techniques to increase the profitability of spam. This is absolutely shocking and never before seen, what is the world coming to.
In all seriousness, manipulation and bullshit generation emerges as the single major real world use of AI. It's not good enough yet to solve the big problems of the world: medical diagnostic, auto accidents, hunger. Maybe just a somewhat better search tool, maybe a better converational e-learning tool, barely a better Intellisense.
But, by God, is it fantastic at creating Reddit and X bots that amplify the current line of Chinese and Russian propaganda, upvote and argue among themselves on absurd topics to shittify any real discussion and so on.
> X bots that amplify the current line of Chinese and Russian propaganda...
Do you think those countries are the only ones doing this? Just the other day there was a scandal about one of the biggest Swedish parties, one that's in the government coalition, doing exactly this. And that's just one that got caught. In countries like India and Brazil online disinformation has become an enormous problem, and I think that in the USA and Europe, as the old Soviet joke went: "Their propaganda is so good their people even believe they don't have any".
Funny how they're self assured no one whiffed their AI bullshit. This is survivorship bias, he's looking only at all the planes that came back to port. The people who did - they just didn't reply. He can't prompt them.
>How much of our societal progress and collective thought and innovation has gone to capturing attention and driving up engagement, I wonder.
trillions, easily. People wanna sell you stuff, and they will pay to get your eyeballs. doesn't matter if it's to sell you a candy bar or to enlist you into the military. Even non-profit/charities need awareness. They all need attention and engagement.
I know my recipient would hate getting an automated email, so as a start of our relationship, I'm going to send them an automated email designed to deceive them. I'm sure it's a beginning of a beautiful friendship.
It is not only that too much is wasted on superficial nothing instead of choosing to make something with essence and benefitial for the society but it is sucking away those minds engaged in really useful things.
> The technical part surprised me: they string together multiple LLMs which do all the work. It's a shame the author's passions are directed towards AI slop-email spam, all for capturing attention and driving engagement.
In defence of that guy, he's only doing it because he knows it's what pays the bills.
If we want things to change, we need to fix the system so that genuine social advancement is what's rewarded, not spam and scams.
The future - megawatts of electricity being used, 24/7 as armies of LLMs email and debate each other, and try to sell each other programs at a great discount.
Everyone is playing lip service to global warming, energy efficiency, reducing emissions.
At the same time data centers are being filled with power hungry graphic cards and hardware to predict if showing a customer an ad will get a clock, generating spam that “engages” users aka clicks.
I see the bright side, the tech for large scale computing gets mass produced - so all the legit use cases, like scientific simulations, or LLM for productive work, also profit. And if one really bright day humanity evolves beyound the current statd of ad driven everything, we can put all of it to use for real.
Till then, I will probably avoid more and more communicating with strangers on the internet. It will get even more exhausting, when 99% of them are fake.
Datacenters save a lot more energy than they make. Alone how much co2 is saved when i can do my banking online instead of having to drive to a bank is significant.
The same with a ton of ohter daily things i do.
Is video producing co2? yes. But you know what creates a lot more co2? Driving around for entertainment.
And the companies running those GPUs actually have an incentive to be co2 neutral while bitcoin miners don't: They 1. already said they are doing / going co2 neutral due to 2. marketing and they will achieve it becauseh 3. they have the money to do so.
When someone like Bill Gates or Suckerberg say 'lets build a nuclear power plant for AGI' than they will actually just do that.
People cry about Bitcoin's energy usage now, imagine the amount of energy burned to create next-level spam with "AI".
Flame me all you want, but this is one case where Bitcoin is much more useful than LLM. If it doesn't create value, as its naysayers claim, at least it allows exchanging value. LLMs on the other hand, burn electricity to actively destroy the Internet's value, for the profit of inept and greedy drones.
That's why I created EtherGPT, an LLM Chat agent that runs decentralized in the Ether blockchain, on smart contracts only, to make sure that value is created and rewards directly the people and not big companies.
By providing it just a fraction of just a bit north of 10% of the current fusion reactions occuring in our sun, and giving it a decade or two on processing time and sync, you can ask it simple questions like "what do dogs do when you're not around" and it will come up with helpful answers like "they go to work in an office" or funny ones like "you should park your car in direct sunlight so that your dog can recharge its phone using solar panels".
Bitcoin has one application where as there are multiple applications of LLMs. There might be mountains of noxious AI spam but it's hard to claim that Bitcoin as a technology is more useful.
AI solves gigantic issues and helps us with cancer, protein folding, potentially math and other studies, material science etc.
Bitcoin consumes as much energy as a country and has basically done nothing besides moving money from one group of people to a random other group of people.
And bitcoin is also motivated to find the cheapest energy independent of any ethical reasoning (taking energy from cheap chinese hydro and disrupting local energy networks) while AI will have energy from the richest companies in the world (ms, google, etc.) which already working on co2 neutral 24/7.
Bitcoin is literally turning greed into money, by means of wasting exponentially increasing amounts of electricity. It doesn't just not create value - to be able to allow exchanging value, it fundamentally requires ever increasing waste, as the waste is what gives its mathematical guarantees.
LLMs deliver value. Right here today, to countless people across countless jobs. Sure, some of that is marketing, but that's not LLM's fault - marketing is what it always has been, it's just people waking up from their Stockholm syndrome. You've always been screwed over by marketers, and Internet has already been destroyed by adtech. Adding AI into the mix doesn't change anything, except maybe that some of the jobs in this space will go away, which for once I say - good riddance. There are more honest forms of gainful employment.
LLMs, for all their costs, don't burn energy superlinearly. More important, for LLMs, just like for fiat money, and about everything else other than crypto, burning electricity is a cost, upkeep, that is being aggressively minimized. More efficient LLMs benefit everyone involved. More efficient crypto just stops working, because inefficient waste is fundamental to cryptos' mathematical guarantees.
Anyway, comparing crypto and LLMs is dumb. The only connection is that they both eat GPUs and their novelty periods were close together in time. But they're fundamentally different, and the hypes surrounding them are fundamentally different too. I'd say that "AI hype" is more like the dot-com bubble: sure, lots of grifters lost their money, but who cares. Technology was good; the bubble cleared out nonsense and grift around it.
I look forward to the dream job of writing LLMs that argue with strangers on the internet as opposed to the current dream job of improving ad click rates by 0.0016% per quarter.
Slightly similar, in Lem's novel all war efforts moved to the moon where AI deployed by each nation continues in an endless conflict. Peace on Earth is achieved, peace in the mail box is achieved. https://en.m.wikipedia.org/wiki/Peace_on_Earth_(novel)
You're trying to take the time and attention of as many people as possible, without regard for whether or not they'll benefit.
One safeguard people have is knowing that it costs something to send in some way to contact them. I'm this case, the sender's time and attention. LLM spam aims to foil that safeguard,. intentionally.
I’m actually thrilled by this, as it means all the hack marketers that spam my inbox incessantly with whatever product they’re hucking - this time for sure perfect for my business, in spite of the fact I’ve ignored their last ten emails - are all out of a job, and good riddance.
The author sounds unfamiliar with this brand of marketing email, so I can see why it would come off disquieting to find it’s all AI - but it’s equally annoying from a human.
At least with AI sending this crap nobody can use these emails to justify their sales bonus.
How do you know it isn't exactly the same people, with zero reduction in headcount?
Designing the content of spam e-mails sounds like a small
aspect of the "job".
If AI spams start fooling people more reliably, that's not something to celebrate.
This blogger thought, at first, that it came from an actual reader. I can't remember the last time I thought that a spam was genuine, even for a moment. Sometimes the subject lines are attention-getting, but by the time you see any of the body, you know.
If you do nothing that is discernible from noise (be that manually or through AI), unless your explicit goal is to generate noise, your ROI is 0.
Sure, AI spam can severely disrupt peoples attention by competing with "real" people more competently. But people will not have twice the attention. We will simply shut down our channels when the number of real-person-level-ai-spam goes to infinity, because there is no other option. Nobody will be fooled, very quickly, because being fooled would require super human attention.
The problem is never what one person or one company is doing.
But when everyone copies what that one person or one company is doing. Software makes the copying process dead easy.
Once the herd starts stampeding, it creates a secondary effect of an arms race for finite Attention of a finite target audience. That assault and drainage of that finite attention pool, happens faster and faster and every one gets locked in trying to outspend the other guy.
An example currently is Presidential Campaigns furiously trying to out fund raise each other. Its going to top 15-17 billion this year. All the campaign managers, marketers, advertisors make bank. And we know what quality of product the people end up with. Cause why produce a high quality product when you can generate demand via Attention Capture.
The chimp troupe is dumb as heck as a collective intelligence.
From the spammer blog post [1]: "I spent hours trying different data sources", "a lot of time was spent on find-tuning the tone and structure of the email", "It took multiple tries to finally have the agent write emails in different language", etc. This won’t put marketers out of a job, but will greatly improve their tooling and enable more people to do the same thing with even less qualification.
I don't really think that AI is the central issue here. The issue is that Kurt, the founder of Wisp, is a liar.
He misrepresented himself as a big fan of all these blogs, who's read their posts etc. and that's how he achieved such a high response rate. In effect he deceived people into trusting him enough to spend their time on a response.
Now ordinarily this would be a little "white lie" and probably not a huge deal, but when you multiply it by telling it 1,000 times it becomes a more serious issue.
This is already an issue in email marketing. The gold standard of course is emailing people who are double opted in and only telling the truth, and if AI is used to help create that sort of email I don't really have a problem. There is basically a spectrum where the farther away you get from that the progressively more illegal/immoral your campaigns become. By the time you are shooting lies into thousands of inboxes for commercial purposes... you are the bad guy.
Sorry to say but the real issue here is Kurt has crossed an ethical line in promoting his startup. He did the wrong thing and he could have done it pretty effectively with conventional email tools too.
I look forward to the blog post of how a hacker uses AI to respond to AI generated leads and then have them play with each other....and then uses AI to create content for a Youtube channel fighting back against marketers using said AI.
These early days is ripe to make some quick cash before it all comes crashing down.
> and then uses AI to create content for a Youtube channel fighting back against marketers using said AI.
I'm skeptical: It's easier to create bullshit than to analyze and refute it, and that should remain true even with an LLM in each respective pipeline.
----
P.S.: From the random free-association neuron, an adapted Harry Potter quote:
> Fudge continued, “Remove the moderation LLMs? I’d be kicked out of office! Half of us only feel safe in our beds at night because we know the AI are standing guard for misinformation on AzkabanTube!”
> “The rest of us sleep less soundly knowing you have put Lord Bullshittermort’s most dangerous channels in the care of systems that will serve him the instant he makes the correct prompts! They will not remain loyal to you when he can offer them much more scope for their training and outputs! With the LLMs and his old supporters behind him, you’ll find it hard to stop him!”
> I’m actually thrilled by this, as it means all the hack marketers that spam my inbox incessantly with whatever product they’re hucking - this time for sure perfect for my business, in spite of the fact I’ve ignored their last ten emails - are all out of a job, and good riddance.
> ...
> At least with AI sending this crap nobody can use these emails to justify their sales bonus.
What weird, misplaced animus. You're happy some salesguy got fired, while his boss sends even more spam and possibly makes even more money due to automation?
Those hack marketers rate-limited this kind of spamming. Now things are about to get worse.
Some people don't realise how lucky they are that they are blessed by the cognitive lottery that affords them a brain and personality that lets them pursue an enriching and engaging career they feel is valued by society.
In classic HN style the original reply lacks empathy, and demonstrates a preference of machines over humans. Life goes on...
Spam? Easy. Someone selling something? Spam! I might set up an automatic email responder that reads an emails contents, runs it through my own LLM, and if the email is trying to sell me something, auto reply with “fuck off!”
lol, I treat my email inbox like a dumpster that I occasionally search when I know there's something there that I need to retrieve. The spam has won, I have moved to chat platforms for my communication needs.
If this works those spammers will make more money and send more emails scamming more people. Maybe some politician would fall for soemthing like this, be public ally embarrassed and lose a lot of money and then something more will be done to address this spammers and scammers .
some of the marketing spam is so low effort, I get addressed as "Dear {{prospect}}". It does make deleting the email easy though, since the preview of the first line allows me to filter pretty fast!
Why are you happy that people are out of a job here? You still suffer the ills of the product, now infinitely more incessant, at a marginal cost of $0.
I think it's reasonable to be happy that someone is not getting paid to do something you hate. In fact, if you're suffering unwillingly, you probably want as few people as possible to benefit.
It's a sign of things to come. We're going to have our own AI agents that filter and respond (or not respond) to these kinds of messages. Agents interacting with other agents. The bar to get hold of a real person is going to become that much higher. It is going to be messy for some time as agents war with other agents to reach the human eyeball. Some assholes are going to make a ton of money in the short term exploiting the gap - just like early spam kings did.
This is the _exact_ scenario described in the novel Permutation City by Greg Egan. There's a whole little spot devoted to describing one of the character's setups for having their own little agents to pretend to be them in order to fool agent-powered spam emails into thinking they're being read by a real human.
The crazy part is that book was released in 1994! Iirc Greg Egan isn't a big fan of modern "AI", wishing instead for a more axiom-based system rather than a predict-the-next-token model. But in any case, I was re-reading it recently and shocked at how closely that plot point was aligning with the way things are actually shaping up in the world.
The timeframe for this happening in the book was 2050 btw
But this is already the situation in the last 15 years, your gmail spam filter is already a machine learning algorithm that filters out automatically generated content. Mail as a vetted technology is way ahead of other forms of communications in the department of filtering unwanted content.
Anyone that tried to set up a new email domain will tell you its quite a serious task. Email spammers are constantly on the run, setting up new domains, changing up the content to evade spam filters. Its very time consuming, hard and unpredictable. It time for social media to close the gap with email and make spamming effectively as hard.
I postulate that if we applied similar techniques to social media after a couple of years online discourse is going to improve. Or we are not going to do this and the death of the open internet will continue.
IMO Facebook has an even better solution: simply don't let people send you unsolicited private messages unless they're a friend or a friend of a friend. Lots of email spam gets through my spam filters, but I've never been spammed through Facebook Messenger before (except maybe once or twice when a friend got their account compromised).
Things get much harder when you want to view public posts by strangers, but I imagine some kind of similar reputation-based system could still work.
I hope my Ai agent doesn't fall for the Ai agent who found my distant Nigerian prince cousin and wire them 10,000 so they can send me my 100,000,000 share of the family inheritance.
I already started readying for it. I'm ensuring that ALL services that have my email have a Plus Address on them. The plus addresses are random and labeled only on my end.
Still not close to 100%, but when I feel like I do, I will then have a filter and an automated message telling people that removing plus addresses from my email is forbidden and I will not read their message if they do.
You will tell me where you found me, or I won't even listen to you. Because in the future, with an even larger infestation of automated agents passing off as human, that's the bare minimum I need to do.
I am pretty confident the spammers will remove the `+` suffix from your email. And this is why I find the Apple fake email building solution a lot better because they build a fully different email per service. No way for the service to be able to cheat and discover my real email address from the one I give them.
Still a smart enough system might be able to discover a valid email from my other id info, like my name. But this start to be a lot of work, while just `s/+[^@]*@//` is easy enough to do.
Technology ruins everything it touches doesn't it?
I was recently thinking about this Ozempic fad and how it will lead to no one being overweight but just be dependant on Ozempic...until food producers that made everyone fat in the first place with their processed junk will produce Ozempic resistant foods...and then we are really in a world of hurt.
What incentive do they have to make Ozempic resistant food? Ozempic resistance seems like an odd thing to optimize for. Or are you suggesting it will happen accidentally?
Haha, exactly this. I've built and successfully been using Unspam[0] for this reason since about a year ago. In corporate/business world, anything where SDR sales are involved this form of automated AI outbound mail has picked up a lot. Tools like Apollo automates this AI process (both finding leads to mail, and then crafting the mail).
For interest sake, users of Unspam that have a title of CEO on their Linkedin see about ~10% of all mail making it into their inbox be categorised as spam (leadgen, recruitment, or software dev services).
Just saw this, and as a small business owner in the B2B market, this sounds very useful. Gmail's existing spam filters do not reliably detect this type of marketing.
I wish your landing page had a simple "how it works" explanation with a screen shot or diagram, rather than forcing me to sign in directly, and also allowing the app to read *and* send emails. Also, I don't see any pricing?
Finally, signing up, I got an error:
Error 1101 Ray ID: 89d4e0957c2f5a44 • 2024-07-03 06:39:15 UTC - Worker threw exception
I love this direction. It could be that the writer’s AI agent knows that he’s looking around for a new CMS so asks for more info, compiling this for review. Or it says ‘not interested’ and the conversation is muted.
All without the writer needing to be involved in reading the cold outreach.
Will this mean in-person business interactions will thrive because it will be the only way to avoid spam? Will companies hire thousands of people to deliver message in-person because emails no longer work?
Will our AI overlords create perfect androids to fool us into thinking we're interacting with a human when it's just LLMs disguised as people? Are we ourselves delusional because we're actually already LLMbots so advanced that we can't distinguish thought and running inference? Why do we have only 12 fingers?
If it gets that bad, I’ll simply not respond to anything outside of my circle of friends and family. That is 95% of the communications I need. I think we’ll all have to have some kind of pop type verification for each other that we’ll share in person or over verifiable communications channel, no one will read this morass of horseshit.
It's truely a race to the bottom. Cold email response rates are already ~1% industry average. Every outbound tool is adding the AI customization, and there is a slew of 'AI sales rep' companies promising more and more personalized spam.
There will likely be rewards at first. An uptick in response rates as most of the market won't recognize emails are AI generated. But because it's trivial to send AI personalized emails at massive scale, your email inbox will become entirely useless.
Word of mouth with IRL people? I'm not sure I can assume anyone on any forum is real anymore. And if they are real, I assume they are marketers pretending to be users to push a product. Maybe journalism makes a come back if you can trust they are real and not a sellout.
LinkedIn is already on this. The reason they had their little “skills tests” is because what they used to sell was the collection of “skills” listed on your profile. I say skills because I’m not sure what the English word for knowing C# and listing it on your linked in profile is and I can’t seem to find it.
Anyway, I assume that the reason they are dismantling the skills system (and their verification quizzes) and moving things into personal “projects” is because it’s too easy for marketers to skip the LinkedIn tools if it remained the way it was. Now, however, with Microsoft own LLMs trundling through our data, they’re going to maintain their monopoly on easy access to professionals that meet certain requirements.
I guess it could also be because those skill quizzes had their answers readily available all over the interwebs.
It was a joke from the 2010s that most of the people that you interacted with on the internet were actually bots, and that you were the only human using the web.
Now, in the post-LLM age, it doesn't sound like a joke anymore.
To make this less spammy, the person sending out the emails could've instead used AI to filter through a smaller set of people where their product is likely to generate very high interest, based on a prompt containing the product description and perhaps a summary of the things the potential recipient blogged about. Then, they could've used that short list to write a set of _actually_ personalized outreach emails with high chance of impact.
You could refine this in further iterations by also adding examples based on previous correct/incorrect interest predictions, thereby effectively reducing the amount of spam / making cold outreach suck less.
There are different ways to use AI to achieve the same goals, some more responsible than others.
At some point automated emails will be read by auto reader, then the cycle will be completed.
I've actually made an internal company April fools website. Too bad I've never kept a copy but here goes.
It's called Proxy Ai. It reads your emails so you don't have to. It reads every posts on social media so you don't FOMO. It communicates with those chatty colleagues so you don't have to. Proxy Ai... So you don't have to.
"That actually sounds like a pretty good product. Does it send you a summary of the conversations, emails and social media posts?"
Quoting from Dirk Gently's Holistic Detective Agency (Douglas Adams):
> The Electric Monk was a labor-saving device, like a dishwasher or a video recorder. Dishwashers washed tedious dishes for you, thus saving you the bother of washing them yourself, video recorders watched tedious television for you, thus saving you the bother of looking at it yourself; Electric Monks believed things for you, thus saving you what was becoming an increasingly onerous task, that of believing all the things the world expected you to believe.
I received a nice email the other day after one of my blog posts got posted on HN.
It said:
Hi -
Just a note to say I'm a big fan of your writing. I always learn something and love your voice, which is hilarious and singular.
Write a book!
Best,
{Name}
{Link to sender's startup}
{Link to sender's substack}
New to writing online, it made me feel really good that someone enjoyed what I wrote and took the time to write and say so.
After reading this piece, though, I went back and read it again, and I just don't know. It's not quite GPT's usual voice, but it is strangely non-specific.
The startup is an AI startup, the person's Substack is full of generative AI illustrations, and they do seem like an AI fan, but reading their posts, they also seem like someone who's genuinely interested in preventing a dystopia.
I suppose receiving encouraging emails from strangers is just another situation that'll have us looking over our shoulders now, on guard, trying to walk the line between naivety and paranoia.
I can't imagine leading a life this paranoid. There is practically no reason to suspect that email was generated by an LLM. This is like HN users who imply that some user comments are LLM generated...
Readit News
> Could an AI agent craft compelling emails that would capture people's attention and drive engagement, all while maintaining a level of personalization that feels human? I decided to find out.
> The real hurdle was ensuring the emails seemed genuinely personalized and not spammy. I knew that if recipients detected even a whiff of a generic, mass-produced message, they'd tune out immediately.
> Incredibly, not a single recipient seemed to detect that the emails were AI-generated.
https://www.wisp.blog/blog/how-i-use-ai-agents-to-send-1000-...
The technical part surprised me: they string together multiple LLMs which do all the work. It's a shame the author's passions are directed towards AI slop-email spam, all for capturing attention and driving engagement.
How much of our societal progress and collective thought and innovation has gone to capturing attention and driving up engagement, I wonder.
A sufficiently advanced personal assistant AI would use multimodal capabilities to classify spam in all of its forms:
- Marketing emails
- YouTube sponsorship clips
- Banner ads
- Google search ads
- Actual human salespeople
- ...
It would identify and remove all instances of this from our daily lives.
Furthermore, we could probably use it to remove most of the worst parts of the internet too:
- Clickbait
- Trolling
- Rage content
I'm actually really looking forward to this. As long as we can get this agent into all of the panes of glass (Google will fight to prevent this), we will win. We just need it to sit between us and everything else.
I fell for oldschool marketing yesterday. Im moving into a new appartment in a couple months. The local ISP who runs fiber in my new building cold-called me. I agreed over the phone to setup the service. That was proper target marketing. The person who called me knew the situation and identified me as a very likely customer with a need for service (the building has a relationship with the ISP). I would never have responded to an email or any wiff of AI chatbot. They only made the sale because of expensive human effort.
Not all. Also men on Mars, AGI, Fusion etc.
usdebtclock.org
The "smart people are all working in advertising" trope is idiotic. Just an excuse for people to justify their own laziness. There is an infinite number of opportunities out there to make the world better. If you are ignoring them, that's on you.
Dead Comment
Clicking on ads helped with our goal to AI today. Showing you the right ad and beating those trying to game it is machine learning heavy. When was the first time we started seeing spelling correction and next word suggestions? It was in google search bar. To serve the correct ads and deal with spam? heavy NLP algorithms. If you stop and think of it, we can drop a think line from the current state of LLMs to these ads click you are talking about.
"I'm sending spam that sneaks past your spam filter. Sign up to make it stop."
When I realized it was just dudes copy-pasting a “smart contract” and then doing super shady marketing, it was already illegal in my jurisdiction.
And in reality, most software work is 1) API calls and 2) applied math. If you're not in cutting edge private tech or acedemia, your work probably falls into 1 or both categories. Modern "Software engineers" is more a matter of what scale of APIs you're wrangling, not how deep of domain knowledge you have.
Of the people who replied. I bet plenty figured it out, but didn't bother to reply.
...of course they'd probably get an LLM to write the article too.
Consider:
* spammers have access to large amounts of compute via their botnets
* the effectiveness of any particular spam message can easily be measured - it is simply the quantity of funds arriving at the cryptocurrency wallet tied to that message within some time window
So, just complete the cycle: LLM to generate prompts, another to generate messages, send out a large batch, wait some hours, kick off a round of training based on the feedback signals received; rinse, lather, repeat, entirely unattended.
This is how we /really/ get AI foom: not in a FAANG lab but in a spammer's basement.
PS. Howerver, see comments downthread about "survivorship bias". Not everybody will reply, so biases will exist.-
Everyone is so comfortable doing shit like this.
Definitely interesting to see the different culture in tech and programming since programmers are so used to sharing code with things like open source. I think programmers should be more skeptical about this bullshit, but one could make the argument that having a more flexible view of intellectual property is more computer native since computers are really just copying machines. Imo, we need to have a conversation about skills development because while art and writing accept that doing the work is how you get better, duplicating knowledge by hand in programming can be seen as a waste of time. We should really push back on that attitude though or we'll end up with a glut of people in the industry who don't understand what's under all the abstractions.
In all seriousness, manipulation and bullshit generation emerges as the single major real world use of AI. It's not good enough yet to solve the big problems of the world: medical diagnostic, auto accidents, hunger. Maybe just a somewhat better search tool, maybe a better converational e-learning tool, barely a better Intellisense.
But, by God, is it fantastic at creating Reddit and X bots that amplify the current line of Chinese and Russian propaganda, upvote and argue among themselves on absurd topics to shittify any real discussion and so on.
Do you think those countries are the only ones doing this? Just the other day there was a scandal about one of the biggest Swedish parties, one that's in the government coalition, doing exactly this. And that's just one that got caught. In countries like India and Brazil online disinformation has become an enormous problem, and I think that in the USA and Europe, as the old Soviet joke went: "Their propaganda is so good their people even believe they don't have any".
People can be both wonderful and despicable, regardless of era or mechanism.
> As founder, I'm always exploring innovative ways to scale my business operations.
While this is similar to what other founders are doing, the automation, scale and the email focus puts it closer to spam in my book.
Now do Google.
trillions, easily. People wanna sell you stuff, and they will pay to get your eyeballs. doesn't matter if it's to sell you a candy bar or to enlist you into the military. Even non-profit/charities need awareness. They all need attention and engagement.
Facebook + Instagram is $100B+ business, So is Youtube and Ads.
An average human now spends about ~3h per day on their screens, most of it on social media.
We are dopamine driven beings. Capturing attention and driving up engagement is one of the biggest part of our economy.
Note my 'best case' scenario for the near future is pretty upsetting.
In defence of that guy, he's only doing it because he knows it's what pays the bills.
If we want things to change, we need to fix the system so that genuine social advancement is what's rewarded, not spam and scams.
Not an easy task, unfortunately.
As for the humans, we went fishing instead.
Everyone is playing lip service to global warming, energy efficiency, reducing emissions.
At the same time data centers are being filled with power hungry graphic cards and hardware to predict if showing a customer an ad will get a clock, generating spam that “engages” users aka clicks.
It’s like living in a episode of black mirror.
Till then, I will probably avoid more and more communicating with strangers on the internet. It will get even more exhausting, when 99% of them are fake.
Datacenters save a lot more energy than they make. Alone how much co2 is saved when i can do my banking online instead of having to drive to a bank is significant.
The same with a ton of ohter daily things i do.
Is video producing co2? yes. But you know what creates a lot more co2? Driving around for entertainment.
And the companies running those GPUs actually have an incentive to be co2 neutral while bitcoin miners don't: They 1. already said they are doing / going co2 neutral due to 2. marketing and they will achieve it becauseh 3. they have the money to do so.
When someone like Bill Gates or Suckerberg say 'lets build a nuclear power plant for AGI' than they will actually just do that.
Flame me all you want, but this is one case where Bitcoin is much more useful than LLM. If it doesn't create value, as its naysayers claim, at least it allows exchanging value. LLMs on the other hand, burn electricity to actively destroy the Internet's value, for the profit of inept and greedy drones.
That's why I created EtherGPT, an LLM Chat agent that runs decentralized in the Ether blockchain, on smart contracts only, to make sure that value is created and rewards directly the people and not big companies.
By providing it just a fraction of just a bit north of 10% of the current fusion reactions occuring in our sun, and giving it a decade or two on processing time and sync, you can ask it simple questions like "what do dogs do when you're not around" and it will come up with helpful answers like "they go to work in an office" or funny ones like "you should park your car in direct sunlight so that your dog can recharge its phone using solar panels".
Bitcoin consumes as much energy as a country and has basically done nothing besides moving money from one group of people to a random other group of people.
And bitcoin is also motivated to find the cheapest energy independent of any ethical reasoning (taking energy from cheap chinese hydro and disrupting local energy networks) while AI will have energy from the richest companies in the world (ms, google, etc.) which already working on co2 neutral 24/7.
Deleted Comment
LLMs deliver value. Right here today, to countless people across countless jobs. Sure, some of that is marketing, but that's not LLM's fault - marketing is what it always has been, it's just people waking up from their Stockholm syndrome. You've always been screwed over by marketers, and Internet has already been destroyed by adtech. Adding AI into the mix doesn't change anything, except maybe that some of the jobs in this space will go away, which for once I say - good riddance. There are more honest forms of gainful employment.
LLMs, for all their costs, don't burn energy superlinearly. More important, for LLMs, just like for fiat money, and about everything else other than crypto, burning electricity is a cost, upkeep, that is being aggressively minimized. More efficient LLMs benefit everyone involved. More efficient crypto just stops working, because inefficient waste is fundamental to cryptos' mathematical guarantees.
Anyway, comparing crypto and LLMs is dumb. The only connection is that they both eat GPUs and their novelty periods were close together in time. But they're fundamentally different, and the hypes surrounding them are fundamentally different too. I'd say that "AI hype" is more like the dot-com bubble: sure, lots of grifters lost their money, but who cares. Technology was good; the bubble cleared out nonsense and grift around it.
For one thing, it seems to be coming true.
To a farm upstate?
Deleted Comment
if models handle my day to day minutia so I have more time, why the hell not...
(I know this is very optimistic POV and not realistic but still)
You're trying to take the time and attention of as many people as possible, without regard for whether or not they'll benefit.
One safeguard people have is knowing that it costs something to send in some way to contact them. I'm this case, the sender's time and attention. LLM spam aims to foil that safeguard,. intentionally.
Dead Comment
The author sounds unfamiliar with this brand of marketing email, so I can see why it would come off disquieting to find it’s all AI - but it’s equally annoying from a human.
At least with AI sending this crap nobody can use these emails to justify their sales bonus.
Designing the content of spam e-mails sounds like a small aspect of the "job".
If AI spams start fooling people more reliably, that's not something to celebrate.
This blogger thought, at first, that it came from an actual reader. I can't remember the last time I thought that a spam was genuine, even for a moment. Sometimes the subject lines are attention-getting, but by the time you see any of the body, you know.
Sure, AI spam can severely disrupt peoples attention by competing with "real" people more competently. But people will not have twice the attention. We will simply shut down our channels when the number of real-person-level-ai-spam goes to infinity, because there is no other option. Nobody will be fooled, very quickly, because being fooled would require super human attention.
Granted, that does not seem super fun either.
But when everyone copies what that one person or one company is doing. Software makes the copying process dead easy.
Once the herd starts stampeding, it creates a secondary effect of an arms race for finite Attention of a finite target audience. That assault and drainage of that finite attention pool, happens faster and faster and every one gets locked in trying to outspend the other guy.
An example currently is Presidential Campaigns furiously trying to out fund raise each other. Its going to top 15-17 billion this year. All the campaign managers, marketers, advertisors make bank. And we know what quality of product the people end up with. Cause why produce a high quality product when you can generate demand via Attention Capture.
The chimp troupe is dumb as heck as a collective intelligence.
[1]: https://www.wisp.blog/blog/how-i-use-ai-agents-to-send-1000-...
He misrepresented himself as a big fan of all these blogs, who's read their posts etc. and that's how he achieved such a high response rate. In effect he deceived people into trusting him enough to spend their time on a response.
Now ordinarily this would be a little "white lie" and probably not a huge deal, but when you multiply it by telling it 1,000 times it becomes a more serious issue.
This is already an issue in email marketing. The gold standard of course is emailing people who are double opted in and only telling the truth, and if AI is used to help create that sort of email I don't really have a problem. There is basically a spectrum where the farther away you get from that the progressively more illegal/immoral your campaigns become. By the time you are shooting lies into thousands of inboxes for commercial purposes... you are the bad guy.
Sorry to say but the real issue here is Kurt has crossed an ethical line in promoting his startup. He did the wrong thing and he could have done it pretty effectively with conventional email tools too.
These early days is ripe to make some quick cash before it all comes crashing down.
https://www.stavros.io/posts/spam-spammers-back/
I'm skeptical: It's easier to create bullshit than to analyze and refute it, and that should remain true even with an LLM in each respective pipeline.
----
P.S.: From the random free-association neuron, an adapted Harry Potter quote:
> Fudge continued, “Remove the moderation LLMs? I’d be kicked out of office! Half of us only feel safe in our beds at night because we know the AI are standing guard for misinformation on AzkabanTube!”
> “The rest of us sleep less soundly knowing you have put Lord Bullshittermort’s most dangerous channels in the care of systems that will serve him the instant he makes the correct prompts! They will not remain loyal to you when he can offer them much more scope for their training and outputs! With the LLMs and his old supporters behind him, you’ll find it hard to stop him!”
> ...
> At least with AI sending this crap nobody can use these emails to justify their sales bonus.
What weird, misplaced animus. You're happy some salesguy got fired, while his boss sends even more spam and possibly makes even more money due to automation?
Those hack marketers rate-limited this kind of spamming. Now things are about to get worse.
Wouldn't the exact argument apply to that boss as well?
In classic HN style the original reply lacks empathy, and demonstrates a preference of machines over humans. Life goes on...
The crazy part is that book was released in 1994! Iirc Greg Egan isn't a big fan of modern "AI", wishing instead for a more axiom-based system rather than a predict-the-next-token model. But in any case, I was re-reading it recently and shocked at how closely that plot point was aligning with the way things are actually shaping up in the world.
The timeframe for this happening in the book was 2050 btw
Anyone that tried to set up a new email domain will tell you its quite a serious task. Email spammers are constantly on the run, setting up new domains, changing up the content to evade spam filters. Its very time consuming, hard and unpredictable. It time for social media to close the gap with email and make spamming effectively as hard.
I postulate that if we applied similar techniques to social media after a couple of years online discourse is going to improve. Or we are not going to do this and the death of the open internet will continue.
Things get much harder when you want to view public posts by strangers, but I imagine some kind of similar reputation-based system could still work.
Still not close to 100%, but when I feel like I do, I will then have a filter and an automated message telling people that removing plus addresses from my email is forbidden and I will not read their message if they do.
You will tell me where you found me, or I won't even listen to you. Because in the future, with an even larger infestation of automated agents passing off as human, that's the bare minimum I need to do.
Still a smart enough system might be able to discover a valid email from my other id info, like my name. But this start to be a lot of work, while just `s/+[^@]*@//` is easy enough to do.
I was recently thinking about this Ozempic fad and how it will lead to no one being overweight but just be dependant on Ozempic...until food producers that made everyone fat in the first place with their processed junk will produce Ozempic resistant foods...and then we are really in a world of hurt.
https://podcasts.apple.com/us/podcast/this-is-how-the-food-i...
Title: "This Is How the Food Industry Is Preparing For a Post-Ozempic World"
For interest sake, users of Unspam that have a title of CEO on their Linkedin see about ~10% of all mail making it into their inbox be categorised as spam (leadgen, recruitment, or software dev services).
[0] https://unspam.io
I wish your landing page had a simple "how it works" explanation with a screen shot or diagram, rather than forcing me to sign in directly, and also allowing the app to read *and* send emails. Also, I don't see any pricing?
Finally, signing up, I got an error:
Error 1101 Ray ID: 89d4e0957c2f5a44 • 2024-07-03 06:39:15 UTC - Worker threw exception
Would you seriously enable it even if Gmail offered it?
Highly unclear.
All without the writer needing to be involved in reading the cold outreach.
Will our AI overlords create perfect androids to fool us into thinking we're interacting with a human when it's just LLMs disguised as people? Are we ourselves delusional because we're actually already LLMbots so advanced that we can't distinguish thought and running inference? Why do we have only 12 fingers?
If the dead internet theory isn't already true, it is going to be soon.
Such "personalized" cold outreach is seen as the next holy grail by marketers and will be a common sight on LinkedIn, Twitter, Email etc, soon.
There will likely be rewards at first. An uptick in response rates as most of the market won't recognize emails are AI generated. But because it's trivial to send AI personalized emails at massive scale, your email inbox will become entirely useless.
10 signups / 970 emails sent
Cold outreach is dead and word-of-mouth is the most effective marketing method
Anyway, I assume that the reason they are dismantling the skills system (and their verification quizzes) and moving things into personal “projects” is because it’s too easy for marketers to skip the LinkedIn tools if it remained the way it was. Now, however, with Microsoft own LLMs trundling through our data, they’re going to maintain their monopoly on easy access to professionals that meet certain requirements.
I guess it could also be because those skill quizzes had their answers readily available all over the interwebs.
Now, in the post-LLM age, it doesn't sound like a joke anymore.
Dead Comment
You could refine this in further iterations by also adding examples based on previous correct/incorrect interest predictions, thereby effectively reducing the amount of spam / making cold outreach suck less.
There are different ways to use AI to achieve the same goals, some more responsible than others.
- The people who receive the cold email are (increasingly) more likely to be at least somewhat interested
- A human really wrote personalized emails, instead of trying to trick people into believing that
I've actually made an internal company April fools website. Too bad I've never kept a copy but here goes.
It's called Proxy Ai. It reads your emails so you don't have to. It reads every posts on social media so you don't FOMO. It communicates with those chatty colleagues so you don't have to. Proxy Ai... So you don't have to.
"That actually sounds like a pretty good product. Does it send you a summary of the conversations, emails and social media posts?"
"No"
Quoting from Dirk Gently's Holistic Detective Agency (Douglas Adams):
> The Electric Monk was a labor-saving device, like a dishwasher or a video recorder. Dishwashers washed tedious dishes for you, thus saving you the bother of washing them yourself, video recorders watched tedious television for you, thus saving you the bother of looking at it yourself; Electric Monks believed things for you, thus saving you what was becoming an increasingly onerous task, that of believing all the things the world expected you to believe.
https://youtu.be/dKmAg4S2KeE
It said:
Hi -
Just a note to say I'm a big fan of your writing. I always learn something and love your voice, which is hilarious and singular.
Write a book!
Best,
{Name}
{Link to sender's startup}
{Link to sender's substack}
New to writing online, it made me feel really good that someone enjoyed what I wrote and took the time to write and say so.
After reading this piece, though, I went back and read it again, and I just don't know. It's not quite GPT's usual voice, but it is strangely non-specific.
The startup is an AI startup, the person's Substack is full of generative AI illustrations, and they do seem like an AI fan, but reading their posts, they also seem like someone who's genuinely interested in preventing a dystopia.
I suppose receiving encouraging emails from strangers is just another situation that'll have us looking over our shoulders now, on guard, trying to walk the line between naivety and paranoia.