AI Search Tool ChatGBT is Woke, Leftist, Gay and Tranny

[Badger]

I've played with ChatGPT enough to determine it's woke, leftist, gay and tranny. They have trained it to be.

I tried an inquiry with it, asking it what is the best alternative theory on the assassination of JFK. It replied back with heavy reliance on the official Warren Report findings. It failed to give a single instance of a well documented alternative theory, even though there is a huge literature on the topic. It kept harping repeatedly on the total lack of evidence for any alternative theory, implicitly stating nothing else is worth mentioning, even though I explicitly asked what those theories were. I tried to push it further, with more "dialoging," but it came back with almost identical Warren Commission pap. It's garbage. You are much better off skillfully using non-AI search engines, such as Swiss Cows or even the notorious Google.

My conclusion is these text-based search AI engines are nothing more than another tool to scam millions of clueless, naive and gullible normies into accepting such tools can be depended upon to give them the final, objective and authoritative word on any and all subject matter. Completely and totally untrue.

Where Does ChatGPT Fall on the Political Compass?

Where does ChatGPT fall on the political compass?

We asked the hot new artificial intelligence system to take four popular political quizzes. Guess what we found?

reason.com

 

[yerrag]

Thanks for trying it out. But I doubt you can expect anything but gaslighting feeding the AI.

AI can be good just as the search engines started out giving us good search results only to be dumbed down and then over time it became a tool of narrative control.

Ditto Wikipedia.

Amazon reviews used to allow comments. Not anymore.

If it improves our lives, it will be defeatured. The truth will set you free, and. freedom is antithetical to civilization.

 

[ThinPicking]

Excellent thread title, shrewd analysis and well drawn conclusion.

 

[charlie]

 

[Badger]

Thanks for trying it out. But I doubt you can expect anything but gaslighting feeding the AI.

AI can be good just as the search engines started out giving us good search results only to be dumbed down and then over time it became a tool of narrative control.

Ditto Wikipedia.

Amazon reviews used to allow comments. Not anymore.

If it improves our lives, it will be defeatured. The truth will set you free, and. freedom is antithetical to civilization.

Well said! A keeper...

 

[Badger]

Thanks for trying it out. But I doubt you can expect anything but gaslighting feeding the AI.

AI can be good just as the search engines started out giving us good search results only to be dumbed down and then over time it became a tool of narrative control.

Ditto Wikipedia.

Amazon reviews used to allow comments. Not anymore.

If it improves our lives, it will be defeatured. The truth will set you free, and. freedom is antithetical to civilization.

"If it improves our lives, it will be defeatured."

This immediately makes me think there is evil or something evil resident to the tech sector. A bad sign.

 

[Badger]

More evidence FagGBT sucks Woke d*ck:

View: https://twitter.com/AlexEpstein/status/1606347326624215040?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1606347326624215040%7Ctwgr%5E%7Ctwcon%5Es1_&ref_url=

@AlexEpstein

Alarm: ChatGPT by @OpenAI now *expressly prohibits arguments for fossil fuels*. (It used to offer them.) Not only that, it excludes nuclear energy from its counter-suggestions.


View: https://twitter.com/AlexEpstein/status/1606347326624215040

 

[ThinPicking]

View: https://twitter.com/elonmusk/status/1606642155346612229?s=20&t=EhJZVsbEvH1OxZpMkNAMhQ


There's a bigger, preexisting danger in having a critical mass of people dumb enough to believe a lie. That's exactly how "AI" (another lie) could frequently pass a turing test, which concerns the competency of the observer and not the program.

Funnily enough, Mr Epstein's (lol) request itself contains a lie. Hydrocarbon fuels are not "fossil fuels". So he already fits the bill.

 

[Badger]

Somebody got Trump's voice and used AI to make a fake recording. Sounds identical to Trump, astonishing. He talks about he and other plutocrats having sex with girls on Epstein's island, and when done with them, Epstein sacrifices them to some pagan diety. Caution, lots of profanity here:


View: https://vocaroo.com/188adfpKIXaz

 

[LadyRae]

Somebody got Trump's voice and used AI to make a fake recording. Sounds identical to Trump, astonishing. He talks about he and other plutocrats having sex with girls on Epstein's island, and when done with them, Epstein sacrifices them to some pagan diety. Caution, lots of profanity here:


View: https://vocaroo.com/188adfpKIXaz

Have you watched any Black Mirror episodes on Netflix?

 

[ThinPicking]

Somebody got Trump's voice and used AI to make a fake recording. Sounds identical to Trump, astonishing. He talks about he and other plutocrats having sex with girls on Epstein's island, and when done with them, Epstein sacrifices them to some pagan diety. Caution, lots of profanity here:


View: https://vocaroo.com/188adfpKIXaz

What's happening with all these statistical inferencing developments is quite fascinating. It's becoming increasingly trivial to generate or imitate text, sounds, images and motion from seeding information.

At the same time we have antithetical tools appearing to detect these outputs. Like these examples. A keen eye and ear can also do this, but the tools can help prove it to another.

GPT-2 Output Detector

GitHub - vidhig/deepfake-image-detection: Predict if an image is real or generated by GANs (fake)

Predict if an image is real or generated by GANs (fake) - GitHub - vidhig/deepfake-image-detection: Predict if an image is real or generated by GANs (fake)

github.com

GitHub - afsalashyana/FakeImageDetection: Fake Image Detection Using Machine Learning

Fake Image Detection Using Machine Learning. Contribute to afsalashyana/FakeImageDetection development by creating an account on GitHub.

github.com

Meanwhile the masses have an aversion to graduating from a "user" to become technically inclined (which everyone can). And as I work a technical career, I don't blame them. So regardless of our ability to spot fakes, along with the "security" situation, I'm a bit concerned this SI wild west is compounding already fertile ground for heavy regulation, surveillance and ID verified internet access.

Tech's day of reckoning is probably coming.

 

[Vinny]

I've played with ChatGPT enough to determine it's woke, leftist, gay and tranny. They have trained it to be.

I tried an inquiry with it, asking it what is the best alternative theory on the assassination of JFK. It replied back with heavy reliance on the official Warren Report findings. It failed to give a single instance of a well documented alternative theory, even though there is a huge literature on the topic. It kept harping repeatedly on the total lack of evidence for any alternative theory, implicitly stating nothing else is worth mentioning, even though I explicitly asked what those theories were. I tried to push it further, with more "dialoging," but it came back with almost identical Warren Commission pap. It's garbage. You are much better off skillfully using non-AI search engines, such as Swiss Cows or even the notorious Google.

My conclusion is these text-based search AI engines are nothing more than another tool to scam millions of clueless, naive and gullible normies into accepting such tools can be depended upon to give them the final, objective and authoritative word on any and all subject matter. Completely and totally untrue.

Where Does ChatGPT Fall on the Political Compass?

Where does ChatGPT fall on the political compass?

We asked the hot new artificial intelligence system to take four popular political quizzes. Guess what we found?

reason.com

 

[Badger]

Have you watched any Black Mirror episodes on Netflix?

No.

 

[Beatrix_]

View: https://twitter.com/echo_chamberz/status/1620133001379348481

 

[Inaut]

I'm beginning to wonder if this AI stuff has been in use for a long time now. I look at all the info coming out (alternative history) and I wonder if it's been used to deceive the masses for far longer than we realize. (ie-historical images, documents etc)

 

[ThinPicking]

I'm beginning to wonder if this *SI stuff has been in use for a long time now. I look at all the info coming out (alternative history) and I wonder if it's been used to deceive the masses for far longer than we realize. (ie-historical images, documents etc)

It's unlikely to have been in wide scale use before there was specialised hardware available to churn particularly difficult mathematical problems involved in model formation ("training") and inference execution. However the underlying concept, that seeding information can be observed, recorded and used to create models to predict, generate, imitate and/or affect that which was observed is probably as old as time. I'll stop short of calling that a feature of the human condition, it's probably more accurate to describe it as a feature of civilisation.

"This is the working hour, We are paid by those who learn by our mistakes"

I reckon the key (21st century) hardware milestones are...

- 2005, major CPU manufacturers begin shifting their design and fabrication focus towards parallelism and multi-core architectures.
- 2006, widescale availability of graphics processing hardware with a "Unified Shader Model". Software frameworks for general purpose massively parallel computation begin appearing.
- Ongoing, continual developments around system integration and heterogeneous computing.
- 2016, widescale availability of hardware capable of executing "tensor" operations in a clock or two.
- Developers off to the races.
- 2020, global "chip shortage". I have a strong suspicion fabrication resources were basically hijacked by the corptocracy/technocracy. They were stockpiling for the next leap.

I'm probably missing something there. And there's more to be said about how open source development efforts are geared towards particular problems at particular times.

 

[Inaut]

It's unlikely to have been in wide scale use before there was specialised hardware available to churn particularly difficult mathematical problems involved in model formation ("training") and inference execution. However the underlying concept, that seeding information can be observed, recorded and used to create models to predict, generate, imitate and/or affect that which was observed is probably as old as time. I'll stop short of calling that a feature of the human condition, it's probably more accurate to describe it as a feature of civilisation.

"This is the working hour, We are paid by those who learn by our mistakes"

I reckon the key (21st century) hardware milestones are...

- 2005, major CPU manufacturers begin shifting their design and fabrication focus towards parallelism and multi-core architectures.
- 2006, widescale availability of graphics processing hardware with a "Unified Shader Model". Software frameworks for general purpose massively parallel computation begin appearing.
- Ongoing, continual developments around system integration and heterogeneous computing.
- 2016, widescale availability of hardware capable of executing "tensor" operations in a clock or two.
- Developers off to the races.
- 2020, global "chip shortage". I have a strong suspicion fabrication resources were basically hijacked by the corptocracy/technocracy. They were stockpiling for the next leap.

I'm probably missing something there. And there's more to be said about how open source development efforts are geared towards particular problems at particular times.

So you don’t think the military developed this technology long before it was released to the public? Akin to almost every other “break through” we experience in all other technologies? I’m not a techie so I’m just axin’

 

[ThinPicking]

So you don’t think the military developed this technology long before it was released to the public? Akin to almost every other “break through” we experience in all other technologies? I’m not a techie so I’m just axin’

Crude proofs of concept, probably. But they had no way to do it in time and at scale. They coopted us for that. They always do.

 

[Regina]

View: https://twitter.com/LeighWolf/status/1620744921241251842

 

[Badger]

ChatGPT does have some value. Asked to come up with a Dr Seuss type poem on trans-personage identity, the AI pumped out a hilarious poem. This is gold.