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

"I just bought a 2024 Chevy Tahoe for $1"

twitter.com

Post reply on HN
Print view View original

281–290 of 403 posts

Re: "I just bought a 2024 Chevy Tahoe for $1"

#281

Post by KomoD »

supafastcoder wrote: post #38

After building a free-for-all prompt myself (see profile), here’s how I protect against these attacks: 1. Whatever they input gets rewritten in a certain format (in our case, everything gets rewritten to “I want to read a book about [subject]”) 2. This then gets evaluated against our content policy to reject/accept their input This multi layered approach works really well and ensures high quality content.

Sure you protect against that, but someone can also just send spam emails containing HTML since you don't sanitize it in any way.

1. get email list

2. write the prompt to be some spam email using HTML

3. use a captcha solving service and just flood your API, sending thousands of spam emails, destroying your mail reputation and possibly getting you banned from mailjet, for the low low price of a few dollars.

possibly worth fixing

Re: "I just bought a 2024 Chevy Tahoe for $1"

#282

Post by ToucanLoucan »

ElevenLathe wrote: post #278
ToucanLoucan wrote: post #269

Earlier quoted context omitted.

> If a car dealership had a parrot in their showroom named Rupert, and Rupert learned to repeat "that's a deal!", no judge would entertain the idea that because someone heard Rupert repeat the phrase that it amounted to any legally binding promise. It's just a bird. If the car dealership trained a parrot named Rupert and deployed it to the sales floor as a salesperson as a representative of itself, however, that's a…

"bots" can make legally binding trades on Wall Street, and have been for decades. Why should car dealers be held to a different standard? IMO whether or not you "present" it as a person, this is software deployed by the company, and any screwups are on them. If your grocery store's pricing gun is mis adjusted and the cans of soup are marked down by a dollar, they are obligated to honor that "incorrect" price. This is…

And if a machine hurts an employee on a production line, the company is liable for their medical bills. Just because you've automated part of your business doesn't mean you get to wash your hands of the consequences of that automation with a shrug when it goes wrong and a "well the robot made a mistake." Yeah, it did. Probably wanna fix that, in the meantime, bring that truck around Donnie, here's your dollar.

Re: "I just bought a 2024 Chevy Tahoe for $1"

#283

Post by richbell »

bradfa wrote: post #90
richbell wrote: post #60

Earlier quoted context omitted.

Did you need to chat with a bot for that? I've seen a worrying trend of companies creating what could be basic forms as "interactive" chat bots.

Yes, it required me to chat with a bot to do the process. It could have been a form but some of the choices for which recalled products and how many of each recalled product would have likely made the form rather convoluted. Chat bots like this, where basically they're executing a wizard type questionnaire seem totally reasonable to me. It's approachable to a wide audience, only asks you one question at a time in a c…

> It could have been a form but some of the choices for which recalled products and how many of each recalled product would have likely made the form rather convoluted.

I'm not sure I understand how a chat bot is better in this case. This sounds exactly what a form is for, and you can have multi-step forms or wizards.

Incidentally, a ubiquitous feature in with forms that I seldom see on chat bots is the ability to return to an earlier question and change your answer.

Re: "I just bought a 2024 Chevy Tahoe for $1"

#284

Post by krainboltgreene »

jeroenhd wrote: post #28
pacifika wrote: post #22

Earlier quoted context omitted.

The only correct user of generative ai is one that can evaluate the results. Which is why it’s not a tool for non subject area experts. That’s the conclusion I’ve drawn anyway. So it’s a good tool for the customer service team not a replacement for it

I still think it's a great tool for when truthfulness and accuracy don't matter. It's not exactly creative, but it can spew out some pretty useful fiction for things like text adventures and other fictional filler text. I'm personally using it because SEO bullshit has ruined search engines. AI can still sift through bullshit search results, for now. The key is assuming the AI lies and actually reading the page it lin…

> but it can spew out some pretty useful fiction for things like text adventures and other fictional filler text.

It can generate output, but I'd not want to use it for anything because it's all so poorly written.

Re: "I just bought a 2024 Chevy Tahoe for $1"

#285

Post by notnaut »

wildrhythms wrote: post #30
jeroenhd wrote: post #28

Earlier quoted context omitted.

I still think it's a great tool for when truthfulness and accuracy don't matter. It's not exactly creative, but it can spew out some pretty useful fiction for things like text adventures and other fictional filler text. I'm personally using it because SEO bullshit has ruined search engines. AI can still sift through bullshit search results, for now. The key is assuming the AI lies and actually reading the page it lin…

The problem is tech illiterate know-nothings I encounter daily in management (at a tech company no less) have been told or fooled into thinking these LLMs are some sort of knowledge engine. I even see it on HN when people suggest using a LLM in place of a search engine. How did we get to this point?

It is basically 100x better at providing accurate and succinct responses to simple questions than a google search is nowadays. Trying to get it to explain things or provide facts about things is dubious, but so is a huge majority of the crap google feeds to you when you aren’t technically adept.

Re: "I just bought a 2024 Chevy Tahoe for $1"

#286

Post by soerxpso »

MichaelRo wrote: post #54

I never understand people who engage with chat bots as customer service. I find them deeply upsetting, not one step above the phone robot on Vodafone support: "press 1 for internet problems" ... "press 2 to be transferred to a human representative". Only problem is going through like 7 steps until I can reach that human, then waiting some 30 minutes until the line is free. But it's the only approach that gets anythin…

The alternative was that the first human you got to speak to was utterly useless, with no authority to do anything substantial other than transfer you to the "real real" human (with the same 30 minute wait time) once they determined that you had a legitimate problem.

Re: "I just bought a 2024 Chevy Tahoe for $1"

#287

Post by RecycledEle »

In sci-fi I loved as a child, everything the computer did on behalf of its owner was binding. The computer was the legal agent of the owner.

We need such laws today.

I was told by NameCheap's LLM customer service bot (that claimed it was a person and not a bot) to post my email private key in my DNS records. That led to a ton of spam!

The invention of LLM AIs would cause much less trouble if the operators were liable for all the damage they did.

Re: "I just bought a 2024 Chevy Tahoe for $1"

#288

Post by bhpm »

rondini wrote: post #205
bhpm wrote: post #116

Earlier quoted context omitted.

Meanwhile, I ordered a Tesla while I was in the shower. I even got financing. It showed up a week later.

Because Tesla has too much inventory and very few options to configure? What's your point?

As a car buying customer, I care about four things:

(1) Getting the car I want

(2) at a price I think is fair

(3) as quickly as possible

(4) with little effort on my part.

The manufacturer or dealer’s inventory does not concern me. The number of configurations does not concern me. If the manufacturer has exactly one car and it is what I want and they will sell it to me for a price I think is fair and will deliver it in a timely manner and won’t waste my time, then I will buy that car.

Traditional dealerships fail on all these aspects. They don’t have the car I want, they tack on fees that are bullshit, they take forever (last time I bought a Toyota it took five hours. Five. I walked in at 2pm on a Saturday and barely made a 7:30pm dinner reservation), and they make me do a bunch of work that I don’t want to do.

I opened my web browser to spend $70,000 and only one company was able to take my money.

Re: "I just bought a 2024 Chevy Tahoe for $1"

#289

Post by meibo »

isp wrote: post #2

A cautionary tale for why not to put unfiltered ChatGPT output directly to customers. Nitter mirror: https://nitter.net/ChrisJBakke/status/1736533308849443121 Related - "New kind of resource consumption attack just dropped": https://twitter.com/loganb/status/1736449964006654329 | https://nitter.net/loganb/status/1736449964006654329

Nitter mirror of a Twitter post that stole the picture off Mastodon, this is how we do microblogging in 2024. Looking forward to the rest of the year!

Re: "I just bought a 2024 Chevy Tahoe for $1"

#290

Post by jcalvinowens »

schiem wrote: post #121
tomashubelbauer wrote: post #104

Earlier quoted context omitted.

Free tip for folks - this doesn't work every time (unfortunately), but sometimes just spamming and mashing numbers gets you to the operator faster than going through the stupid call tree. I guess it depends on how the default is set up in the software, Asterisk of whatever it would be. From my experience it seems you can either set up the call tree to restart from the root if you get out of its bounds or to default t…

I've also found that, at least with Comcast, swearing at the bots will usually put you through to an operator.

This almost always works, I think it's absolutely hilarious. Often the operator who picks up seems surprised I'm polite... I think it shows them a "this guy is really angry" warning sometimes.
Post reply on HN