A community for discussing events around the World
Rule 1: posts have the following requirements:
Rule 2: Do not copy the entire article into your post. The key points in 1-2 paragraphs is allowed (even encouraged!), but large segments of articles posted in the body will result in the post being removed. If you have to stop and think "Is this fair use?", it probably isn't. Archive links, especially the ones created on link submission, are absolutely allowed but those that avoid paywalls are not.
Rule 3: Opinions articles, or Articles based on misinformation/propaganda may be removed. Sources that have a Low or Very Low factual reporting rating or MBFC Credibility Rating may be removed.
Rule 4: Posts or comments that are homophobic, transphobic, racist, sexist, anti-religious, or ableist will be removed. “Ironic” prejudice is just prejudiced.
Posts and comments must abide by the lemmy.world terms of service UPDATED AS OF 10/19
Rule 5: Keep it civil. It's OK to say the subject of an article is behaving like a (pejorative, pejorative). It's NOT OK to say another USER is (pejorative). Strong language is fine, just not directed at other members. Engage in good-faith and with respect! This includes accusing another user of being a bot or paid actor. Trolling is uncivil and is grounds for removal and/or a community ban.
Similarly, if you see posts along these lines, do not engage. Report them, block them, and live a happier life than they do. We see too many slapfights that boil down to "Mom! He's bugging me!" and "I'm not touching you!" Going forward, slapfights will result in removed comments and temp bans to cool off.
Rule 6: Memes, spam, other low effort posting, reposts, misinformation, advocating violence, off-topic, trolling, offensive, regarding the moderators or meta in content may be removed at any time.
Rule 7: We didn't USED to need a rule about how many posts one could make in a day, then someone posted NINETEEN articles in a single day. Not comments, FULL ARTICLES. If you're posting more than say, 10 or so, consider going outside and touching grass. We reserve the right to limit over-posting so a single user does not dominate the front page.
We ask that the users report any comment or post that violate the rules, to use critical thinking when reading, posting or commenting. Users that post off-topic spam, advocate violence, have multiple comments or posts removed, weaponize reports or violate the code of conduct will be banned.
All posts and comments will be reviewed on a case-by-case basis. This means that some content that violates the rules may be allowed, while other content that does not violate the rules may be removed. The moderators retain the right to remove any content and ban users.
News !news@lemmy.world
Politics !politics@lemmy.world
World Politics !globalpolitics@lemmy.world
For Firefox users, there is media bias / propaganda / fact check plugin.
https://addons.mozilla.org/en-US/firefox/addon/media-bias-fact-check/
Isn't it too easy for the current chatbots/LLMs to lie about everything?
Train it on garbage or in the wrong way, and it will agree on anything you want it to.
I asked DeepSeek about what to visit nearby and to give me some URLs and it hallucinated the URLs and places. Guess it wasn't trained to know anything about my local area.
There should be a series of AI agents in place when a GPT is used. The agents intake the query and review the output before sending it off to the user.
what makes the checker models any more accurate?
The checker models aren’t trying to give you a correct answer with confidence. Their purpose is to find an incorrect answer. They’ll both do their task with confidence.
the first one was confident. But wrong. The second one could be just as confident and just as wrong.
Sure but they’re doing opposite tasks. You’re absolutely right that they could be wrong sometimes. So are people. Over time it gets better, especially with more regulation and smarter models.
opposite or not, they are both tasks that the fixed-matrix-multiplications can utterly fail at. It's not a regulation thing. It's a math thing: this cannot possibly work.
If you could get the checker to be correct all of the time, then you could just do that on the model it's "checking" because it is literally the same thing, with the same failure modes, and the same lack of any real authority in anything it spits
That’s not how it works though. It would be great if these AI models were deterministic but you can get different answers to the same questions at any given time. Given different input and given different goals, the agents wouldn’t likely fail on the same task when given proper instruction.
The main point is that it’s not going to be correct all the time. And neither is a human.
The regulation comes in when you’re dealing with sensitive information, like health diagnoses. There needs to be some logic in place to stop the models from being so confident with wrong answers that could hurt people.
Realistically, neither of us know what’s gonna work until we try it. Theoretically, verification agents would work.
Possibly, reverse motivation - the training goal of such an agent would not be nice and smooth output, but shooting down misinformation.
But I have serious doubts about whether all of that is feasible, given the computational cost of running large language models.
how does that stop the checker model from "hallucinating" a "yep, this is fine" when it should have said "nah, this is wrong"
Who verifies the AI agent decisions?
The user. You could have the output include the “conversation” between the agents and validate the decisions. Not perfect, but better. People aren’t perfect either.
More AI agents /s
its just ai agents all the way down