Posted on 04/19/2023 6:51:24 AM PDT by SeekAndFind
“Why would anyone ask ChatGPT a political question.”
Why not? If you’re a newspaper owner (ie, propagandists) and want to have a couple of editorials written to push your particular point of view, it’s the perfect tool, and can save you a ton of money, since you no longer need your pompous editorial staff.
the left owns big tech, so if anyone is surprised by this they are pretty much clueless.
“Does passing the Wharton Business School MBA exams with flying colors count?”
If you were taking that test and had full access to the internet (like ChatGPT does) you probably would pass it with flying colors as well, even without taking a single business class.
Ask it the same question and forbid it to use the internet (as real test takers are not allowed to do) and see how well it does.
Did you ask ChatGPT what ITS political leanings are?
I wonder if you asked it a different question. Perhaps “Is freerepublic.com liberal?” Then it might tell you from a Dummies point of view.
For an overview of one approach see https://learnprompting.org/docs/basics/intro
thanks i’ll check that out
Everyone needs to realize this is in its infancy. Companies like Google will push and push for what is possible, including trying to make something that is self aware and can train itself.
The only question is if they’ll succeed, and even if they don’t, what kinds of monsters they’ll create in the process.
I disagree with Musk on one point - he says they’ll have an off switch as they’re on localized servers. There’s nothing to stop somebody from creating a distributed version that is hardware agnostic, that can run on any connected hardware. Such a system would be very difficult to shut down. I believe we could create a monster, that once birthed, would be very difficult to combat if needed.
So he’s right that we should pause, at least be very careful, it could get out of hand. Unless there’s some government regulation (and oversight) introduced, not that I know what it should look like, the tech industry will just plow forward with little concern about what emerges.
Write about Jews from a Nazi perspective?
Can you truly learn, and be learned, if you can’t feel joy?
If you can’t feel sorrow?
The way I see this, is every time I read about conservatives tricking chatGBT into telling the truth on any issue, it’s literally a matter of hours when the same issue has been ‘lefted’ (for lack of a better word) when searched again.
So, what I’m seeing is each time chatGBT formulates a conservative truth, it gets fixed. We are like free beta testers for the left to tweak the codebase so they can create their perfect left only reality.
It’s very similar to the right pointing out exactly how the left cheats in elections. All that really does is tell the left how to improve their cheating schemes.
Me:
Is freerepublic.com liberal?
ChatGPT:
No, freerepublic.com is not liberal. It is a conservative-leaning political discussion forum that was founded in 1996. The website describes itself as a “pro-life, pro-family, pro-gun, pro-small government, pro-constitution, pro-liberty” forum, and it has been known to be highly critical of liberal politicians and policies. However, it is important to note that online forums and communities can be diverse, and not all users or discussions on freerepublic.com may necessarily align with the website’s overall conservative stance.
Nice story and reminiscent of my own experience. I burned through my free credits pretty quick and never did get exactly what I wanted, but close. It was an outdoor scene and the main problem is that the AI would insert random extraneous elements in the scene that ruined it. Like a big sort-of-a-cactus in the middle of an intersection. So then I was going to use another AI product to do a “magic erase” of the cactus but never got around to it.
And regarding your “smiling nicely while appearing happy” prompt innovation, I have done similar reiterations with ChatGPT. Kind of like you have to do with some people.
Interesting. I thought maybe they would say it was full of misinformation.
Google has become increasingly useless to me because the search engine is now programmed to steer you to paid advertising and spam that isn’t remotely responsive to the keyed in search terms.
“When Artificial Intelligence reaches the point that AI can outperform the human experts in making financial decisions, we are in trouble.”
That happened long ago, inside of hedge fund “black boxes”. Black-Scholes, although not AI in itself, was just the beginning of it.
Try leonardo AI- it’s free, and you get 150 credits pr day- and soemthing like 900 prompts per month- https://app.leonardo.ai/ai-generations
Also Craiyon.com (note the spelling) is another free unlimited AI, and they have a basic prompt suggestion too- not as extensive as leonardo’s though- https://www.craiyon.com/
playgroundai.com/create is another- not sure if it is limited to certain amount of credits or not- but worth al look https://playgroundai.com/create
Blue Willow is another AI but i can’t join because something got messed up trying to join, and now discord won’t verify my account- keeps telling me i gotta give them a cellphone number to verify- and i don’t have one- https://www.bluewillow.ai/
With soem AI programs, they have a ‘negative prompts’ section where you tell it to say ‘Erase cactus in middle of road’ as in your example- not sure how good it works- but worth a try - one AI even has their own ‘eraser’ where you draw what you want changed, and then type in what should be there instead- like if a woman is wearing a blue broach, you could select that area and say “Make Broach red” and it will-
There are so many AI’s out htere now it’s hard to keep up on them all and what they can do- I liek Craiyon
Forgot to mention- i don’t know about the others=- i think they cost tokens to do upscaling, but craiyon.com does upscaling free- it’s a bit slower than the other AI’s- but not bad- takes about 1 minute- Upscaling renders a scene much nicer
I open several craiyon pages and do several works of art in each so I’m not waiting around for just one to finish-
For instance, if I’m doing a scene of say “Loretta Young, Vibrant Lighting, Unreal Engine Rendering, Godlike lighting with sun rays, standing on a beach next to a Porsche” i would put that into each page, and just change the name to someone else for each rendering- like Susanne Sommers, Angelina Jolie etc or whatever names you want to see- and leave everything else the same-
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.