The danger of people using and accepting chatGPT style AI this whole time is that it has never been effective, never been ethical, and now it’s going to be used by the government to harm people at scale

Comments

You can really see how the "move fast and break things" ethos has now infested DC. It's why they are rushing this out years before it's ready, trying to use it for stuff it might never be suitable for, and especially why they want to dump off the labor of getting it working.
Lots of DeepSeek on this app.
I think most normal people just cant comprehend how 'funny AI images' makes the leap to 'AI is determing who gets sent to the gulag'

I think a real issue at the moment is all of AI being thought of as falling under the same umbrella.
the silly thing is that most folks who work in or studied computer science (at least that I've interacted with so maybe a skewed sample) also think that current LLMs are overhyped. it really feels like a capitalist class fantasy exclusively most of the time
And I don’t want to hear the “how could this happen” chorus when people have been extremely anti-AI fearing these outcomes the WHOLE TIME!
reply
15 years ago, I worked at UPenn on AI research on event detection in videos. What, last year, OpenAI released ChatGPT? And now we're using it in government. That timeframe is terrifying. It's not sufficiently tested to be a consumer product let alone something the federal government relies on.
And by "ChatGPT" I mean any of the LLMs that have been released in the last year or so, in case that's not clear.
I mean, how many Hollywood movies, books, and TV shows have there been instilling a fear of AI, and it seems none of that took root?
I'll admit though that I can't think of any media example where the AI was dumb or randomly wrong? Maybe that's the problem? We've been condition to fear a pervasive, all-knowing AI that would turn malicious, but instead we are facing a pervasive, B.S.-ing AI that we havent been taught is dangerous?
Mildly interesting how the time bw “the crazy-sounding people warning us were actually right” keeps getting shorter.

Like, with AI we are at maybe at about one year?
You have to constantly fight inertia that connivence can give you in these situations.
I worked for a lot of people who were working on a lot of these teams during 2020-2022 and they were having full on existential crisis while they spent all their working hours on teams or zoom or whatever on efforts to spin the narrative of the technology into a story of “this will help locate
missing people and reunite families” or “this will help people in *developing countries or underprivileged communities* have access to essential tools that will allow them to participate in the global economy..” blahblah
But they were fucking sweating. They were so freaked out.
Depends on the use case. Being able to convert trig & calc to C# functions in Unity with ChatGPT when the typical reply in forums is "dunno, have you tried discord?" and the discord reply is "dunno, have you tried the forums?" makes my life as a solo indiegame developer a lot easier.
And just like that;

Now you're under control, and you do what they told ya, now you're under control and you do what they told ya.
The revolution will not be televised; but it turns out, it will be on social media.
A quick reminder many songs were already on an instaban list when 9/11 occured.
we said that what the machine says is wrong

they said that they will burn down reality until what the machine says is right
The director and another individual in my department uses ChatGPT often and we work in threat and vulnerability management cybersecurity… the funny thing is we aren’t supposed to be using ChatGPT on our work computers due to cybersecurity concerns 💀💀💀
My sister just got a Facebook friend recommendation for a person (foster child) she's only mentioned on her government email.
👆🏽👆🏽👆🏽
I think we are severely ignoring the environmental impact of these super computers, and I'm not even talking about the long term pollution from energy use on the planet. I'm talking direct, waste byproduct, running these things getting in to the air/ground water/etc.

People are getting sick
It’s at-scale harm babyyy!
Like a lot of things, AI is a tool. It is neither good nor bad.
Like everything else in this world, it is the use that matters and on that, I tend to agree w/you - I think too many use it because they're either unwilling to do the hard work or looking for an easy out to success.