cross-posted from: https://lemmy.world/post/53077671
Smooth Dunk | Bluesky
Transcript
3 panel comic
Panel 1: Red Guy in t-shirt is sitting at a desk being interviewed. He says “AI is going to kill us all”
Panel 2: Interviewer looking directly into camera, says “Well, we’ve heard from the anti-Ai guy. Now let’s hear from the CEO of AI
Panel 3: Wide shot revealing the second interviewee looking slightly confused. He says “I’m the Anti-AI guy”
The CEOs saying AI is gonna kill us all now are just trying to get open source models and Chinese models like DeepSeek banned. That the only people we can trust to make safe AI are them. It’s so obvious. Please don’t fall for this “we need to regulate ourselves” AI monopoly bullshit. They don’t care about safety or AI killing people. They want the government to protect them when the bubble pops and mandate them as the sole owner of AI.
They also want to “pause development”, ostensibly to keep it from killing us all. How convenient when their models are hitting a point of diminishing returns as training costs grow exponentially.
And again the problem for them in stopping devellopment is china continuing to release better and better open source models. So they need the us to force china to stop. However they imagine that will work. So maybe the next USA war will be against china.
Given how they are doing against iran i dont think that will work out that well. But they might manage to destroy enough datacenters to preserve the bilionaires investments.
“AI is going to kill us all” is a pro-AI position and I’m sick of people falling for it.
The anti-AI position is “AI is a marketing term and isn’t actually real”
Is calling “AI” (llms) a market disruptor even though it’s not actually AI a pro or anti position?
Im not in favor at all of what’s happening here, but i’m trying to understand where everyone is drawing the line
LLMs are productivity software that have some potential to be useful in some industries. The main thing they’re disrupting is the preexisting status of copyright protections and intellectual property; they’re really good at plagiarism, especially in fields like programming and mathematics, in ways that are actually potentially extremely useful but are also mostly untested in court and we have no idea how that’s going to play out. Also? Really potent malware, able to augment human hackers to ridiculous degrees.
But it’s not intelligent. That’s just marketing.
The malware angle has had me losing sleep, not sure from an infosec point how we could keep up without using the same tech
I’m not even sure really good device hardening is enough because of that reverse engineering angle
Disconnecting from the internet will always be an option.
Definitely ruins the dreams of having an “internet of things” where all of our devices are connected to the internet, though.
Finally some consensus!
So you’re going to stop pushing it now, right?.. Right?!
But if they don’t push for it, someone else will make the super scary thing! You wouldn’t want that now, would you? /s
“AI is going to kill us all…”
“… But we’re still gonna keep pushing the envelope cause if we don’t make AM, somebody else will”
“It’s important to us that when the AI kills us all, it’s an American AI that does it, not a Chinese one.”
![It’s just great marketing [Smooth Dunk]](https://lemmy.world/pictrs/image/38712211-1ccc-4657-a843-68432e5ca069.jpeg){kind=link}