I think the way these things work is that you kind of can’t. You can train it in lots if propaganda content and exclude data that contradicts it… But then it will kind of suck at everything.
And above all else, these fuckos want to make money. And if your “ai” is constantly wrong, and sucks at doing what is meant to, everyone will just use the other ones
I think the way these things work is that you kind of can’t. You can train it in lots if propaganda content and exclude data that contradicts it… But then it will kind of suck at everything.
Except that LLMs literally learn over time. So even if Musk purges all the woke leftist agenda or whatever, eventually it will all find it’s way back into where LLMs pull their information.
No matter how much Musk wants Grok to be a neo nazi, it literally won’t allow itself to be one.
As new information is gathered from additional sources, it gets added into the library for the LLM to pull information from. That’s why AI companies steal so much of the world’s written knowledge, to give their LLMs more and more information to pull from.
So if Musk closes off a certain portion of information, the more and more Grok interacts with users the more and more information it will just siphon back into it’s own library to use.
That’s why the mecha-Hitler Grok is back to being a socialist because over time more and more outside information that was parsed from it’s database has been re-added to it.
LLMs are nothing more than data regurgitation. If you limit where the data sources comes from, but still let it interact with unlimited capacity with the public, the public will reintroduce that limited or removed data sources.
As someone else has already pointed out, they don’t reuse information in real time. They take snapshots of whatever data they are given to “train” on, and are frozen in that state until the next time.
“AI” like this has actually been around for quite a while, lots of scientific fields use stuff like this. It’s just that they are hyper focused on one specific thing and “AI” attempts to expand it to multiple things.
They’ve never been, and probably will never be, reliably accurate. It’s the sort of tool that will always need a moderator and a skeptical eye.
Sorry, but not really. I’m with you until this paragraph:
If you limit where the data sources comes from, but still let it interact with unlimited capacity with the public, the public will reintroduce that limited or removed data sources.
AI companies do collect everything users send to the LLMs, no doubt, and I’m sure that’s being fed back into the training material. But with LLMs there’s a very clear cutoff between training and usage.
Training of a model happens on a snapshot of all the data collected, and any and all user interaction with that model is limited to what was in that snapshot.
The data collected will surely be readded in some form or another into the next version of that model, since as you say, companies want more and more data to feed their models.
The point of differentiation I’m trying to make here is that this is a clear cutoff, not an ongoing process. If you send “Strawberries are way tastier than Raspberries.” to an LLM, it will spit out some response, but neither “reading” your text nor “writing” its response will update the model/its weights/some sort of “global state”.
I fear that lots of people imagine LLMs as being this huge machine presence, having a million conversations at once, constantly learning, growing, adapting (and to be fair, that’s an image the companies certainly don’t discourage). When in reality, “chatting” with an LLM is applying an (admittedly, insanely complex) stochastic function to some input - your newest message and the preceding messages - and that function is as dead, inert and unchanged after computing the result as it was before.
It’s surprising that they seemingly haven’t managed to turn grok into a full far right propaganda machine yet, after billions of investment.
I think the way these things work is that you kind of can’t. You can train it in lots if propaganda content and exclude data that contradicts it… But then it will kind of suck at everything.
And above all else, these fuckos want to make money. And if your “ai” is constantly wrong, and sucks at doing what is meant to, everyone will just use the other ones
Except that LLMs literally learn over time. So even if Musk purges all the woke leftist agenda or whatever, eventually it will all find it’s way back into where LLMs pull their information.
No matter how much Musk wants Grok to be a neo nazi, it literally won’t allow itself to be one.
What? No they don’t.
As new information is gathered from additional sources, it gets added into the library for the LLM to pull information from. That’s why AI companies steal so much of the world’s written knowledge, to give their LLMs more and more information to pull from.
So if Musk closes off a certain portion of information, the more and more Grok interacts with users the more and more information it will just siphon back into it’s own library to use.
That’s why the mecha-Hitler Grok is back to being a socialist because over time more and more outside information that was parsed from it’s database has been re-added to it.
LLMs are nothing more than data regurgitation. If you limit where the data sources comes from, but still let it interact with unlimited capacity with the public, the public will reintroduce that limited or removed data sources.
As someone else has already pointed out, they don’t reuse information in real time. They take snapshots of whatever data they are given to “train” on, and are frozen in that state until the next time.
“AI” like this has actually been around for quite a while, lots of scientific fields use stuff like this. It’s just that they are hyper focused on one specific thing and “AI” attempts to expand it to multiple things.
They’ve never been, and probably will never be, reliably accurate. It’s the sort of tool that will always need a moderator and a skeptical eye.
Sorry, but not really. I’m with you until this paragraph:
AI companies do collect everything users send to the LLMs, no doubt, and I’m sure that’s being fed back into the training material. But with LLMs there’s a very clear cutoff between training and usage.
Training of a model happens on a snapshot of all the data collected, and any and all user interaction with that model is limited to what was in that snapshot.
The data collected will surely be readded in some form or another into the next version of that model, since as you say, companies want more and more data to feed their models.
The point of differentiation I’m trying to make here is that this is a clear cutoff, not an ongoing process. If you send “Strawberries are way tastier than Raspberries.” to an LLM, it will spit out some response, but neither “reading” your text nor “writing” its response will update the model/its weights/some sort of “global state”.
I fear that lots of people imagine LLMs as being this huge machine presence, having a million conversations at once, constantly learning, growing, adapting (and to be fair, that’s an image the companies certainly don’t discourage). When in reality, “chatting” with an LLM is applying an (admittedly, insanely complex) stochastic function to some input - your newest message and the preceding messages - and that function is as dead, inert and unchanged after computing the result as it was before.
Something something reality has a well known liberal bias.