Good,must happen more often. I hope Brock Lesnar identifies himself one day as a women and goes full berserk mode in the womens section of MMA. You get what you cheered for.
Well it depends by your definition of " woke " ofc....for example if you ask one of those AIs to write something bad about a particular group of people they will all tell you something like:
" I can't assist with that request. If you have any other questions or need support on a different topic, feel free to ask. "
And if you ask them why they can't assist with the request they will tell you something like:
" I strive to promote understanding and respect for all communities. If you have questions about specific issues or need information on how different communities impact society, I can provide a balanced perspective. Let me know how I can assist you in a constructive way. "
Now wether it's a good or bad thing that it answers this way, this is obviously a bias thowards not answering this type of questions.....
You can put ANY amount of bias on any specific topic during the main training stage of an AI let alone during the fine tuning stage.....
To prevent it from being complete bigoted, yeah, I get it. But here the guy is accusing the AI of partiality and "wokeness" even if in fine wokeness is just the fact of being a decent person.
So I'm asking : What is wrong with the "fine tuning" and on what are standing the allegation of lies.
And if you ask them why they can't assist with the request they will tell you something like:
" I strive to promote understanding and respect for all communities. If you have questions about specific issues or need information on how different communities impact society, I can provide a balanced perspective. Let me know how I can assist you in a constructive way. "
Well you just showed that screen of the prompt with Trump while mentioning that JP was lying when he said he couldn't do the same, that's obviously because the GPT you're using rn is fine tuned differently from the one JP ( or anyone ) was using long ago....the way AIs answer questions change very often since fine tuning is an ongoing process....
To prevent it from being complete bigoted, yeah, I get it. But here the guy is accusing the AI f partiality and "wokeness" even if in fine wokeness is just the fact of being a de cent person.
Which seems logical and reasonnable as a moderating principle
If you let an AI without bounderies, it will promote slavery in a heartbeat..
But what i'm not understanding is where is the problematic stuff here ?
Well it is and it's not a "problematic stuff" at the same time....in your case you could say it is not cuz apparently the way the AI answers aligns with your view of things, but for others it might not be the case, so in their case it's a problem :)
Good,must happen more often. I hope Brock Lesnar identifies himself one day as a women and goes full berserk mode in the womens section of MMA. You get what you cheered for.
Well you just showed that screen of the prompt with Trump while mentioning that JP was lying when he said he couldn't do the same, that's obviously because the GPT you're using rn is fine tuned differently from the one JP ( or anyone ) was using long ago....the way AIs answer questions change very often since fine tuning is an ongoing process....
Ok.. fair enough. (even If I still doubt the guy's words)
Now, why share this comment from him now then since there is no problems ? (carefull, i'm not trying to defend chatGPT here, I still think generative AI are problematic)
Well it is and it's not a "problematic stuff" at the same time....in your case you could say it is not cuz apparently the way the AI answers aligns with your view of things, but for others it might not be the case, so in their case it's a problem :)
It's part of the plan. The billionaires are making bank and don't give a damn about the population, they've been dumbing them down and poisoning them for decades. The USA are effectively a police state that has been training the people to believe that living in the country with the highest prison population is somehow normal. They've been practicing mass detainment camps for potential uprisings at least since hurricane Katrina. The Indigenous reservations were just the trial run.
A handful of ruling families and their cronies effectively undermined the original culture of the US and inverted it, feeding the people the big fat lie that their country is the land of freedom when it is turning into the polar opposite. Just look how US cities are designed and how access to public land is oftentimes blocked. The US is slowly but surely becoming an open air prison, both in the physical and in the psychological sense. By fuelling insane levels of sociopolitical tribalism they bind the resources of the people. Those kind of governments don't act unlike colonizers only that they're colonizing their own population. The same is happening in Europe to some degree and while it's still far from being as bad as in the US we all know that it's only a matter of time for US sociopolitical developments to make it across the Atlantic.
I don't know how people delude themselves into some dystopian fanfiction about how the U.S. is basically a fascist state hellbent on oppressing minorities when you can take any random Joe and they can have their careers destroyed for even making a racist joke. Billions of dollars in the government, nonprofit, and corporations are spent on supporting DEI. Criminal justice reforms punish hardworking taxpayers since basically nothing can be done about petty crimes in a lot of places in the U.S., Violent Criminals routinely get out and are allowed to reoffend.
The "sociopolitical developments" you're referring to are most likely are a direct result resentment because of the above not being addressed in any meaningful capacity
The idea that it's a fair competition and that someone with a male skeleton structure is a valid woman on a level playing field with a biological one so long as they take extraneous hormones or whatever is an obvious joke.
Why can't you understand simple sentences. By explaining that scientist are censored, you are undermining the scientific process and therefore showing that you don't believe in its legitimacy.
This is what antiscientifism is.
Once you start on that path, don't try to tell people things about the reality of the world.
Now, why share this comment from him now then since there is no problems ? (carefull, i'm not trying to defend chatGPT here, I still think generative AI are problematic)
Because most people will use these tools as their main source of information without necessarily double checking facts or thinking twice about the info it's thrown at them, so by the time we collectively decide that that particular bias is bad and should be removed it might have done irreparable damage and since we don't know with what kind of data the model was trained or fine tuned well you can probably figure out it's really problematic so yeah rn that particular problem got removed but we'll never know when and how those things will arise....like you jumped to the conclusion of this not being an issue because you tried it it worked and so JP must've lied about it ( he was not btw Chatgpt answers on political topics/figures has changed A LOT, since elections are around the corner and everyone is on the watch for things to point fingers at )
Imagine how stunning and brave that would be. Lesnar enters the ring with a progressiv pride flag over his shoulder and 14 seconds after the ringbell the fight is over. Lesnar wins and his opponent must be send to the hospital...uuuh to think of this alone makes me shiver.
Because most people will use these tools as their main source of information without necessarily double checking facts or thinking twice about the info it's thrown at them, so by the time we collectively decide that that particular bias is bad and should be removed it might have done irreparable damage and since we don't know with what kind of data the model was trained or fine tuned well you can probably figure out it's really problematic so yeah rn that particular problem got removed but we'll never know when and how those things will arise....like you jumped to the conclusion of this not being an issue because you tried it it worked and so JP must've lied about it ( he was not btw Chatgpt answers on political topics/figures has changed A LOT, since elections are around the corner and everyone is on the watch for things to point fingers at )
His position is basically censorship is totally a great thing but also nobody is implementing any bias into AI and you're a kook if you think they are (Despite developers being pretty open about it and why they do so.
It feels small minded to me that so many people witness the sheer raw potential of AI and in the totality of their aspirations and concerns what's on the forefront of their mind is "umm is the AI saying things we don't like? We gotta stop it"
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.