It takes effort to be an informed citizen. Artificial intelligence tools offer an alluring shortcut — but they’re not without risk.
It takes effort to be an informed citizen. Artificial intelligence tools offer an alluring shortcut — but they’re not without risk.
Has anyone tested the prompts on these? I’m kind of wondering if truth having a leftist bias could help inform people against the right-wing propaganda.
We can hope that, but their responses rely entirely on where they pull their information from, and I lost all hope that they were actually factual when they turned grok into Mecha Hitler!
Models don’t “care” about “truths”, they are about the average likelyhood that the next word is correct based on the previous x words based on the data they were trained on, which was mostly internet posting.
Their output can also be manipulated by adding “system prompts” so that the user’s question isn’t the only thing being input.
ChatGPT doesn’t think, it doesn’t “reason”, it doesn’t know what is true, it doesn’t “feel” “integrity”. It regurgitates Wikipedia articles against the biases of 4chan and Reddit posts and tries to sell that diarrhea as bowls of soup.
Personally I’m to the point where I understand gulags now. Like, why would they need that? Why would they do something that seems so intentionally cruel and forceful and then read shit like this and I go “oh, oh yeah. I get it now. Yeah some people belong there.”
I very much doubt the kind of people that most need to be in things like gulags end up in them…
This is screaming into the the ether, but I’ll point out that this is not how GPT, or transformer architectures in general, work. You are describing a basic ass bag of words vectorization model, which is certainly in the lineage of transformers but is absolutely not the end of it. What made transformers such an inflection point in NLP is the development of the attention mechanism. It is still generating an iterative and sequential probabilistic series of tokens, but those probabilities are informed by context and attention along a huge scaffold of parameters that encode a lot of semantics. This isn’t exactly how humans reason, but it’s distressingly not that far off either.
That is all to say, this is still a terrible idea but, no, models are not just “hurr durr weighted averages hurr durr” and it’s been over a decade since that was even remotely at the forefront of NLP.
These datasets, along with the rules governing them, are a bit more complicated than weighing 4chan and wiki equally. They absolutely weigh peer-review journals and white-listed encyclopedia sources higher. Once they establish a baseline of credibility and identifying a certain grade-level of communication, along with identifying critical-thinking patterns (fallacies, formal logic, etc.)., then you can quite easily filter out the noise.
Every single time I’ve used a GPT to analyze a conversation with a Trump supporter and myself, it has identified their fallacies I had already identified, and rightfully pointing out fallacies of my own at times – but almost always concludes their argument is far, far worse.
since when do LLMs care about what is true or not? I must’ve missed that, we fixed the hallucination problem?
also, rest assured, Grok has almost certainly been tweaked to say “whoever’s on the Republican ticket”, and probably most other big AI companies in the US who benefit from this administration will do the same, just less ham-fistedly and more subtly than Musk will. If flagship LLMs in general were to start giving ringing endorsements of leftists, I have zero doubt in my mind that would not be allowed to continue, not under the current regime.
But AI has a hallucination problem similar in the way every single human being in existence has an imperfect recollection hallucinatory problem. We have to distinguish accuracy from precision and take that with a grain of salt. I do wonder when the last time people here have used some of these later GPTs, because they absolutely have value and vast consensus of scientists agree.
How one regulates it, is of course another story that I agree with you on. I’ve already used AI successfully to combat far-right trolls though.
I don’t think the truth is going to factor in, but my father has been outsourcing his thinking to Gemini lately and in that case it’s fairly straightforward: they have a profile on him as a moderate Democrat, so it suggests anyone with that tag. It likely isn’t going to fact check that tags, though.
But back to truth: it will pull top search results which very much can be gamed, just like pre-LLM days. That means a false story pushed on a candidate is popular, it’ll default to that and could mislead a voter. That, and of course whatever weights the owner of that LLM sets (although, that is much more subtle than other commenters let on; a weight generally isn’t absolutely, it simply tips things so like, Democratic Socialists won’t magically get recommended MAGA but conservative independents will based on higher values set to conservative tags like immigration views).