• 0 Posts
  • 10 Comments
Joined 1 year ago
cake
Cake day: July 20th, 2023

help-circle


  • Anyone aware if they are also getting data from their slack for government offering? I was looking at the govslack site and I can’t tell one way or the other. While they claim to meet most of the big compliance regs I don’t see anything about training AI being included/excluded.

    I know that stealing trade secrets is a concern but seems like stealing state secrets might have some other implications. I know you’re not supposed to talk on slack about any classified info, but that doesn’t mean that sensitive info isn’t shared which also has some rather profound implications as well.


  • Your response really highlights that you do not get what I’m saying. I’m not arguing it should be banned. I’m saying that acknowledging that the barrier of entry was lowered is at least somewhat of an important factor to consider. Doing it the way flipper did is irresponsible at best, and more realistically ethically corrupt. It’s been done though and you can’t put the cat back in the bag.

    Now governments are trying to ban them, but when 100s of new clones come out I can almost guarantee governments are going to start doing increasingly silly shit to stop it. Do you think that giving every joker a key to any kia/Hyundai is going to lead to governments cracking down on security on the manufacturing side? Or do you think it’ll just give them a bigger excuse to make invasive laws? I’m pretty sure I know where it’ll lead and I seriously doubt it will be leveling laws against the poor old car manufacturers that donate to campaign funds…


  • I’m onboard with that but putting it at the level of operating a tv remote really casts a wider net. You essentially have to be barely literate to use the thing, where before you had to at least be able to read and execute some walkthroughs. Also you had to kind of be in the security/tech scene to even understand that it was an option, where the flipper has, for a lack of a better word, popularized the attack.

    There’s a reason that when you go on sites like exploit db well over half of the exploits require some fiddling to make work. Metasploit is similar as well because it requires you to actually be able to use a cli on some level. While that isn’t a huge bar of entry, it’s still keeps the riff raff out for the most part. The flipper pretty much said fuck it, and let not only the skiddies in, but any dipshit with $80 buy a car stealing autopwn.




  • I don’t disagree with your point, but the flipper zero for sure lowers the bar of entry. Before the flipper came out the, “You must be this tall to ride” required some pretty good knowledge of microcontrollers, hardware peripherals, and software engineering. The people that had that sort of knowledge tended to actually have paying jobs, which is like the biggest factor in not being a street criminal.

    The flipper made the barrier of entry at about the level of being able to operate a TV remote which any dipshit can do. However, the fact that the flipper exists at all means that the cat is out of the bag. As you said, someone else is just going to come along and release a similar product. You can’t just ban the flipper and expect it to have any impact. My concern is they will decided to make certain code illegal, which gets really stupid.


  • It’s escapism I think. At least that’s part of it. Having a machine that won’t judge you, will serve as a perfect echo chamber, and will immediately tell you AN answer can be very appealing to some. I don’t have any data, or any study to back it up, just my experience from seeing it happen.

    I have a friend who I feel like I kind of lost to chatgpt. I think he’s a bit unhappy with where he is in life. He got the good paying job, the house in the suburbs, wife, and 2.5 kids, but didn’t ever think about what was next. Now he’s just a bit lost I think, and somehow convinced himself that people weren’t as good as chatting with a bot.

    It’s weird now. He spends long nights and weekends talking to a machine. He’s constructed elaborate fictional worlds within his chatgpt history. I’ve grown increasingly concerned about him, and his wife clearly is struggling with it. He’s obviously depressed but instead of seeking help or attempting to figure himself out, he turned to a non-feeling, non-judgmental, emotionless tool for answers.

    It’s a struggle to talk to him now. It’s like talking to a cryptobro at peak btc mania. The only thing that he wants to talk about is LLMs. Trying to bring up that maybe spending all your time talking to a machine is a bit unhealthy invokes his ire and he’ll avoid you for several days. Like a herion addict struggling with addiction, even pointing out the obvious flaws in what he’s doing makes him distance himself more from you.

    I’m not young, not old exactly either, but I’ve known him for 25 years in my adult life. We met in college and have been friends ever since. I know many won’t quite understand but knowing someone that long, and remaining close, talk every few days, friends is quite rare. At this point he is my longest held friendship and I feel like I’m losing him to a robot. I’ve lost other friends to addiction in my life and to say that it’s been similar is under stating it. I don’t know what to do for him. I don’t know if there’s really anything I CAN do for him. How do you help someone that doesn’t even think they have a problem?

    I guess my point is, if you find someone who is just depressed enough, just stuck enough, with a particular proclivity towards computers/the internet then you have a perfect canidate for falling down the LLM rabbit hole. It offers them an out to feeling like they’re being judged. They feel like the insanity it spits out is more sane than how they feel now. They think they’re getting somewhere, or at least escaping their current situation. Escapism is very appealing when everything else seems pointless and sort of gray I think. So that’s at least one type of person that can fall down the chapgpt/LLM rabbit hole. I’m sure there’s others out there too with there own unique motivations and reason’s for latching onto LLMs.