malwaretech,

A while back I wrote a couple of blog posts about how I think LLMs will be a net negative (at least in the near term), due to the extreme overhyping giving people unrealistic expectations of their ability. The direct result will be an overall degrade in internet usability as people begin flooding platforms with low quality spam that they erroneously believe to be high quality. Previously it was fairly easy to spot someone who doesn't know what they're talking about, but now LLMs enable them to word things convincingly enough to waste the time of even domain experts.

Even now I'm still often surprised by all the creative ways that people are finding to waste other's time. I just saw this post from one of the curl maintainers reporting that they've been receiving nonsense bug bounty reports based on LLM hallucinations, which I imagine is likely due to people trying to automate bug hunting despite lacking the understanding to confirm their finding. They reported that in one case the submission was convincing enough that they went over the code 3 times before coming to the conclusion that no bug existed and the report was likely AI generate.

https://daniel.haxx.se/blog/2024/01/02/the-i-in-llm-stands-for-intelligence/

Npars01,
@Npars01@mstdn.social avatar

@malwaretech

AI is a blend of the trend towards "Bullshit Jobs" described by David Graeber and the #enshittification business model being forced upon the economy, democracy, & consumers as described by @pluralistic

Rent seeking conduct disguised as "progress".

NovemberMan,

@Npars01 @malwaretech @pluralistic 👍💯👍
Read an article this week that Microsoft will be adding an AI button to new laptops, and already hopping I can buy one without it, it disable it.
When my current one gets obsolete, of course.

Npars01,
@Npars01@mstdn.social avatar
z3z,
@z3z@mastodon.scot avatar

@malwaretech I moderate in a couple of forums and we're already receiving quite a number of LLM generated comments that manage to be... nearly... but not quite on topic.

New member comments are moderated, so these posts are likely made to get the account over the posting threshold so they can unleash proper spam.

It's a nightmare cos you have to carefully read through each one & they're often long. Give me old school Viagra comments! At least you knew instantly what you were dealing with.

malwaretech,

@z3z sometimes it's funny to search reddit for "As of my knowledge cutoff" and see how many people didn't even bother to remove the telltale ChatGPT phrase from their post

maegul,
@maegul@hachyderm.io avatar

@malwaretech @z3z

So what happens to the open internet?

Like search, does the whole thing just become a sort of wild west where you never know if it's worth using/reading?

How much of the internet "battens down" and goes closed?

Do some things return to manual/physical/in-person processes?

I'm certainly less interested in going anything online these days. Spam webpages and shitty search was already enough. This is just the nail in the coffin IMO.

simon_brooke,
@simon_brooke@mastodon.scot avatar

@malwaretech this is pretty much what happened in the #AI boom in the 1980s. The hype was ludicrous, the systems didn't deliver, and for a decade it was impossible to get work as a serious Artificial Intelligence investigator.

People who actually do care about working towards #AI ought to be standing up and saying very publicly that #LLMs contain nothing resembling artificial intelligence at all, because if they don't, we'll see another crash, and another drought.

the5thColumnist,

@malwaretech

LLMs are based on lots of information, but even if it was accurate, which it mostly isn't, information is only knowledge, not intelligence. Intelligence is how you analyze and understand the information. Without that it's just words.
https://the5thc.blogspot.com/2024/01/ai-has-nothing-to-do-with-intelligence.html

pseudonym,
@pseudonym@mastodon.online avatar

@malwaretech

This.

It was already bad a year ago, shortly after chatGPT went public, for librarians flooded with requests for non existent books, with plausible titles, from real authors.

#libraries #books #llm

jab01701mid,

@malwaretech Seems like a new form of a DDoS attack. No actual people could keep up with the rate of incoming fire from AI bots and generators.

jab01701mid,

@malwaretech Captcha on bug bounty submissions ? Can ChatGPT solve Captcha's yet ?

malwaretech,

@jab01701mid they can still AI generate reports and then submit them manually

fbmac,
@fbmac@masto.ai avatar
spacerog,
@spacerog@mastodon.social avatar

@malwaretech I have seen many similar time wasting reports lately.

ed209,

@malwaretech I recently saw an article that was probably AI generated about the Fenix 6, but the really weird part was the photo was also AI generated, of a Fenix 6 that does not, and never will exist. What a time to be alive.

VictimOfSimony,

@malwaretech
The post reads as an argument for more Discord-like platforms where your presence is at the behest of small-scale moderation & the content isn't searchable. Everyone seems to hate that just as much, but the inability for webcrawlers to make the content searchable & profitable off-site seems to be the only complaint anyone has. If you're going to Google to search Reddit you're feeding the problem. Creating inline tools for content creators to make their own contributions searchable & meaningful rewards for them is how you solve this problem.

Once again, if anyone wants to be a part of this solution, I'm still looking for project staff.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • random
  • DreamBathrooms
  • everett
  • InstantRegret
  • magazineikmin
  • thenastyranch
  • rosin
  • GTA5RPClips
  • Durango
  • Youngstown
  • slotface
  • khanakhh
  • kavyap
  • ngwrru68w68
  • ethstaker
  • JUstTest
  • osvaldo12
  • tester
  • cubers
  • cisconetworking
  • mdbf
  • tacticalgear
  • modclub
  • Leos
  • anitta
  • normalnudes
  • megavids
  • provamag3
  • lostlight
  • All magazines