Google's AI is the dumbest one of all. The only thing good about it is that it shows you which sources are powering it's thinking.
Machine Learning and Artificial Intelligence Thread
- Thread starter Johnnyvee
- Start date
Why do you use n_____ in yourr quote but say the full word in the next parargraph?
Just curious.
Because the point of the post is that the kindle version censored it that way, but I was curious if the original edition did. I found out the original was uncensored.
This is from an account that was banned during covid and was recently unbanned by youtube.
He uploads his book warning about AI and digital IDs to Googles new AI software and it makes a podcast discussing his book. It's very convincing.
He also has it clone his voice. It's good at replicating him.
He uploads his book warning about AI and digital IDs to Googles new AI software and it makes a podcast discussing his book. It's very convincing.
He also has it clone his voice. It's good at replicating him.
JR5
Other Christian
Entering the Forcefield: How Language Shapes Reality
This post explores the contrast between two fundamentally different approaches to language and meaning as revealed through large language models. One approach is empirical, consensus-driven, and designed to flatten contradiction for broad readability; the other treats language as a living forcefield of paradox, contradiction, and ecstatic insight, a vehicle capable of shaping perception, thought, and the symbolic architecture of reality. Using a single charged text about the Russia-Ukraine war as a test case, it illustrate how the same prompt may produce radically divergent outputs depending on the epistemic framework chosen.
https://neofeudalreview.substack.com/p/entering-the-forcefield-how-language
This post explores the contrast between two fundamentally different approaches to language and meaning as revealed through large language models. One approach is empirical, consensus-driven, and designed to flatten contradiction for broad readability; the other treats language as a living forcefield of paradox, contradiction, and ecstatic insight, a vehicle capable of shaping perception, thought, and the symbolic architecture of reality. Using a single charged text about the Russia-Ukraine war as a test case, it illustrate how the same prompt may produce radically divergent outputs depending on the epistemic framework chosen.
https://neofeudalreview.substack.com/p/entering-the-forcefield-how-language
I recently posted a tweet with a video of a possum getting surprised by a Halloween decoration. Turns out it was an AI video, which I didn't realize at the time, but it was obvious in hindsight.
AI is tricky in that it can fool people and make then think something was real when it wasn't. Last year Gov Gruesome from CA was mad when an AI video showed him saying something he didn't. There were videos with Kamala Harris like this as well. I thought they were hilarious, but they felt they were the victims of deception.
It was pretty obvious to most people that these were fakes, and that they were satire, but some people probably were fooled. What if a fake video of Trump was made that ended up doing serious political damage? It's easy to imagine situations where dishonest AI content is clearly wrong.
Nowadays there's tons of AI content made for laughs or clicks, like the possum video I posted. People have been calling this stuff AI slop. How should we react to this? Some of it is harmless, but maybe some of it is not. Maybe AI slop is harmful in a way that similar to the harm done by social media.
We're all going to be dealing with AI content from now on. Should everybody learn to recognize AI content and develop an instant prejudice against it? We'll probably all be using it ourselves to make content for work or hobbies or fun.
Should all AI be tagged as such? What about here on the forum? If someone posts a funny video that turns out to be AI, should it be labeled as such?
I think we'll all be having to decide on things like this.
AI is tricky in that it can fool people and make then think something was real when it wasn't. Last year Gov Gruesome from CA was mad when an AI video showed him saying something he didn't. There were videos with Kamala Harris like this as well. I thought they were hilarious, but they felt they were the victims of deception.
It was pretty obvious to most people that these were fakes, and that they were satire, but some people probably were fooled. What if a fake video of Trump was made that ended up doing serious political damage? It's easy to imagine situations where dishonest AI content is clearly wrong.
Nowadays there's tons of AI content made for laughs or clicks, like the possum video I posted. People have been calling this stuff AI slop. How should we react to this? Some of it is harmless, but maybe some of it is not. Maybe AI slop is harmful in a way that similar to the harm done by social media.
We're all going to be dealing with AI content from now on. Should everybody learn to recognize AI content and develop an instant prejudice against it? We'll probably all be using it ourselves to make content for work or hobbies or fun.
Should all AI be tagged as such? What about here on the forum? If someone posts a funny video that turns out to be AI, should it be labeled as such?
I think we'll all be having to decide on things like this.
I believe it's exactly the same as social media. All of these technologies are built with a certain intent that is embeded in it. The technology is not neutral, you have to work against it for good.
It's built for profit and profit is accomplished by your attention.
We are telling our kids not to believe anything unless they see it and experience it in person in real life.
Everything else is to be treated with skepticism.
Note: I'm not trying to promote AI in any way, just my personal opinion
In late September, OpenAI released a new subscription plan called ChatGPT Go. It is much cheaper compared to other plans such as ChatGPT Plus and ChatGPT Pro. I decided it was worth trying since it’s very affordable, and if I didn’t like it, I could simply cancel my subscription the following month. For context, I’ve been a heavy AI user for quite some time — I use both Microsoft Copilot and ChatGPT daily.
I often use ChatGPT to brainstorm ideas or quickly gather an overview of information I need — something that’s accurate enough to give me a good starting point. For example, when I want to learn about the carnivore diet or how intermittent fasting works, I first use ChatGPT to build an initial understanding. Then, I follow up with internet searches and YouTube videos for a deeper dive. For instance, I might watch Shawn Baker’s videos for more insight into the carnivore diet, or Jason Fung and Pradip Jamnadas for in-depth explanations about intermittent fasting. Having ChatGPT give me a quick briefing first makes it much easier to understand these topics later.
For subjects I’m only mildly curious about — the kind I don’t want to research deeply — ChatGPT still performs well. I can ask things like what makes Earth special compared to other planets, why Uranus is colder than Neptune, or what a Hot Jupiter is, and it provides concise, easy-to-understand answers.
I also use ChatGPT to reorganize and clean up YouTube auto-captions. For example, when I watch a long video — say, a two-hour discussion about intermittent fasting and insulin resistance — it’s not practical to rewatch the entire thing just to extract the key points. So, I copy and paste the auto-generated captions into ChatGPT and ask it to arrange them into readable paragraphs. Not only does it structure them properly, but it also corrects grammar and spelling errors from the captions, making the text far easier to read.
Another feature I use constantly is grammar and structure correction for emails and documents. I rely on it so much that I honestly can’t remember the last time I sent an email or printed a document without having ChatGPT review it first.
I’ve also enjoyed experimenting with image generation — not for serious work, but for fun. Both Bing Image Creator/Copilot and ChatGPT can generate images, but ChatGPT has an advantage: it can change the art style of an image, not just create new ones from scratch. As a ChatGPT Go subscriber, I’ve noticed I can create images faster and in greater quantity compared to the free version. So far, I’ve already made over 30 images since subscribing.
Another impressive feature is OCR and text extraction from documents and images. Both Copilot and ChatGPT can do this effectively. In the past, capabilities like these were limited to standalone OCR software such as OmniPage, which came bundled with old Canon scanners. Those older systems often struggled with characters like “rn” and “m,” but modern AI tools like Copilot and ChatGPT can easily interpret them accurately.
My main point is this: AI isn’t inherently evil or a “mark of the beast.” It all depends on how we choose to use it.
In late September, OpenAI released a new subscription plan called ChatGPT Go. It is much cheaper compared to other plans such as ChatGPT Plus and ChatGPT Pro. I decided it was worth trying since it’s very affordable, and if I didn’t like it, I could simply cancel my subscription the following month. For context, I’ve been a heavy AI user for quite some time — I use both Microsoft Copilot and ChatGPT daily.
I often use ChatGPT to brainstorm ideas or quickly gather an overview of information I need — something that’s accurate enough to give me a good starting point. For example, when I want to learn about the carnivore diet or how intermittent fasting works, I first use ChatGPT to build an initial understanding. Then, I follow up with internet searches and YouTube videos for a deeper dive. For instance, I might watch Shawn Baker’s videos for more insight into the carnivore diet, or Jason Fung and Pradip Jamnadas for in-depth explanations about intermittent fasting. Having ChatGPT give me a quick briefing first makes it much easier to understand these topics later.
For subjects I’m only mildly curious about — the kind I don’t want to research deeply — ChatGPT still performs well. I can ask things like what makes Earth special compared to other planets, why Uranus is colder than Neptune, or what a Hot Jupiter is, and it provides concise, easy-to-understand answers.
I also use ChatGPT to reorganize and clean up YouTube auto-captions. For example, when I watch a long video — say, a two-hour discussion about intermittent fasting and insulin resistance — it’s not practical to rewatch the entire thing just to extract the key points. So, I copy and paste the auto-generated captions into ChatGPT and ask it to arrange them into readable paragraphs. Not only does it structure them properly, but it also corrects grammar and spelling errors from the captions, making the text far easier to read.
Another feature I use constantly is grammar and structure correction for emails and documents. I rely on it so much that I honestly can’t remember the last time I sent an email or printed a document without having ChatGPT review it first.
I’ve also enjoyed experimenting with image generation — not for serious work, but for fun. Both Bing Image Creator/Copilot and ChatGPT can generate images, but ChatGPT has an advantage: it can change the art style of an image, not just create new ones from scratch. As a ChatGPT Go subscriber, I’ve noticed I can create images faster and in greater quantity compared to the free version. So far, I’ve already made over 30 images since subscribing.
Another impressive feature is OCR and text extraction from documents and images. Both Copilot and ChatGPT can do this effectively. In the past, capabilities like these were limited to standalone OCR software such as OmniPage, which came bundled with old Canon scanners. Those older systems often struggled with characters like “rn” and “m,” but modern AI tools like Copilot and ChatGPT can easily interpret them accurately.
My main point is this: AI isn’t inherently evil or a “mark of the beast.” It all depends on how we choose to use it.
I recently posted a tweet with a video of a possum getting surprised by a Halloween decoration. Turns out it was an AI video, which I didn't realize at the time, but it was obvious in hindsight.
AI is tricky in that it can fool people and make then think something was real when it wasn't. Last year Gov Gruesome from CA was mad when an AI video showed him saying something he didn't. There were videos with Kamala Harris like this as well. I thought they were hilarious, but they felt they were the victims of deception.
It was pretty obvious to most people that these were fakes, and that they were satire, but some people probably were fooled. What if a fake video of Trump was made that ended up doing serious political damage? It's easy to imagine situations where dishonest AI content is clearly wrong.
Nowadays there's tons of AI content made for laughs or clicks, like the possum video I posted. People have been calling this stuff AI slop. How should we react to this? Some of it is harmless, but maybe some of it is not. Maybe AI slop is harmful in a way that similar to the harm done by social media.
We're all going to be dealing with AI content from now on. Should everybody learn to recognize AI content and develop an instant prejudice against it? We'll probably all be using it ourselves to make content for work or hobbies or fun.
Should all AI be tagged as such? What about here on the forum? If someone posts a funny video that turns out to be AI, should it be labeled as such?
I think we'll all be having to decide on things like this.
Good post, and one that I hope prompts others to think a bit more critically about the convenient, entertaining or otherwise "harmless" uses of AI.
Re: Watching that possum clip -- at first I was like
Then I was like
And then it hit me. . .
AI has now come for our Funny and Cute Animal Videos Thread
My response in one word:
For those interested in breaking down the AI tells of the possum vid, see below
Yes to all of the above. Context follows below.
I also suggest creating a thread called "AI slop dump" where we move posts that include AI videos that either went viral with the masses and/or got past one CiK member (without shaming anyone), along with associated comments.
Further, I believe every CiK member should take steps to educate themselves about how to spot AI so they don't inadvertently participate in its mass uptake.
With this in mind, I encourage all CiK readers to take a moment before posting to consider the credibility and authenticity of the said image, video, tweet, or other data. Some related tips from the same account:
A Few Reasons to be Cautious about AI
Most AI generation is a form of manipulation. Data is created, amended, or removed to produce an unreal version of things. AI bends your perception of reality to fool you into the believing something is real when it is not.
People are instinctually wary of deliberate and clear manipulation, given that it is against the manipulated individuals' interest. Yet AI is becoming so effective that it is bypassing this highly functional defence system. Indeed this defence system is critical to our understanding of reality. It's hard to overstate how important it is to have a clear perception of one's objective environmental conditions.
Yes, some AI videos are funny, especially if we know they are AI generated (example: dishwashing Olympics vid). This all seems relatively harmless and perhaps when presented alone, especially to a high IQ adult with normal cognitive abilities, it would be.
However, AI is and will be largely successful to the extent that it deceives humans, especially those with reduced mental capabilities, including children and the elderly. Ultimately, deception is a form of lying, which is sinful. With this in mind, a strong case can be made that generative AI is inherently dehumanising and antithetic to Christian values.
The mass response to AI so far has largely looked like this:
Some people are waking up though, even if they don't quite know why AI makes them uneasy
Moreover, AI uproots human originality, creation, effort, beauty, and meaning. Why bother learning to draw and create something special and unique, but get frustrated with the slow progress in skills; when you can just type in an AI prompt and get a result in seconds? Meanwhile, transhumanistic technocrats present it to us as fun, convenient, or efficient. And are making billions from controlling it.
I appreciate your post and wanted to like it, because of the practical tools you shared. Clearly, AI tools have many professional uses. Thank you for your contribution.
Nonetheless, this final comment seems to reflect a somewhat naive understanding of the real, dark side of AI. The following issues are disturbing but should be shared for people who are ambivalent about the pros/cons of AI:
The Dark Side of AI: Risks to Children - Child Rescue Coalition
AI is popping up online and on social media, but is there a risk to children? Learn how to protect kids who are using AI.
childrescuecoalition.org
The article continues with some general ideas about what to do in response:
Then there's even more problems:
-> It reduces critical thinking. Article summary:
-> It hinders creativity. Article summary:
-> It creates distance between humans. Young people especially are now using AI bots for social and emotional support.
-> And the consequences of chat bot buddies can be dire:
-> And on and on it goes.
Guys, let's not sleepwalk into this. We still have a sphere of control in our own lives. It is time to increase our vigilance and conscious attention about how we use AI, and how our children, parents, families and friends interact with it.
Related video discussion
Overview:
AI’s Unseen Risks: How Artificial Intelligence Could Harm Future Generations
The Great Simplification #180 with Zak Stein
natehagens.substack.com
What's scary is the time is coming when you won't even be able to discern a real video from AI.