Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I Received an AI Email

timharek.no

Post reply on HN
Print view View original

421–430 of 597 posts

Re: I Received an AI Email

#421

Post by projektfu »

b112 wrote: post #340
whiplash451 wrote: post #329

Earlier quoted context omitted.

No, I really meant structured. Extracting data from structured documents is surprisingly hard when you need very high accuracy. What I mean by structured is: invoices, documents containing tables, etc. Extracting useful data from fully unstructured content is very hard IMO and potentially above the capacity of LLMs (depending on your definition of "useful" and "unstructured")

But this is why I made my complexity statement in my other reply. Why are firms sending around invoices, tables instead of parseable data. Oh I know the argument, because "so hard to cooperate" on standards, etc. Madness.

Partly because the standards, such as X12, have a high startup cost to use them, they aren't very opinionated about the actual content, and you have to get the counterparty on board to use them.

Re: I Received an AI Email

#422

Post by echelon »

madamelic wrote: post #420
echelon wrote: post #415

Earlier quoted context omitted.

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. Fur…

> Enough of us will get pissed off that we will develop AI agents that sit between us and the internet. Until _that_ company gets overrun by MBAs who are profit-driven then they start injecting ads into the results. It will come in the vein of "we are personalizing the output and improving responses by linking you with vendors that will solve your problems".

There will be a uBlock Origin for that.

Re: I Received an AI Email

#423

Post by russnewcomer »

FeepingCreature wrote: post #343
tsukikage wrote: post #320

Earlier quoted context omitted.

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…

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.

That’s one of Peter Watt’s Rifters trilogy, I think maybe the second one? Been a few years since I read them. I think it’s a biological neural net, not an Ai per se. Lots of big ideas in those books, but not a lot of optimism and some rough stuff.

Re: I Received an AI Email

#424

Post by cranberryturkey »

devjab wrote: post #128
kibwen wrote: post #43

Earlier quoted context omitted.

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.

Does it really matter if you’re being cold called by an AI or some sales person following the same few procedures they always do? I’d prefer sales people keep their jobs. Having had the misfortune of being seated next to the telemarketing team in an investment bank for half a year… however… Let’s just say that I’m not sure you would even know if it was a person or a bot. They’re not even scripted or “trained” like yo…

Very true, I sat next to the sales guy at a small company and he was on the phone all day repeating himself day in and day out. Easily replaced by AI.

Re: I Received an AI Email

#425

Post by __loam »

highspeedbus wrote: post #345
ossyrial wrote: post #238

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 spam…

Things I wish become taboo: Admitting to use AI content. Everyone is so comfortable doing shit like this.

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.

Re: I Received an AI Email

#426

Post by joelthelion »

ossyrial wrote: post #238

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 spam…

> 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.

Not an easy task, unfortunately.

Re: I Received an AI Email

#427

Post by yazantapuz »

throwaway0665 wrote: post #193
sph wrote: post #180

Earlier quoted context omitted.

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.

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.

Well, a friend of mine built its house thanks to btc last's ath. Surely someone is cashing out nvidia right now. Indirectly useful :)

Re: I Received an AI Email

#430

Post by internet101010 »

paulluuk wrote: post #416
brabel wrote: post #358

Earlier quoted context omitted.

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.

As someone who actually has a university degree in Artificial Intelligence, I feel like this is always how it's been. Before, an "AI Expert" was someone who knew how to use Tensorflow, PyTorch or Keras. Before that, an "AI Expert" was someone who knew how to write a Monte Carlo simulation, etc etc. You could of course say the same for frontend engineer or backend engineers. How many frontend engineers are simply impo…

I would liken it to cars. There is a difference between engineers, mechanics, and mechanics that know a certain car so well that they fabricate parts that improve upon the original design.
Post reply on HN