The CIA doesn't want the pubic to know how long advanced AI technology has been around
When a tech is “just released”
That means its been in use behind the scenes for far longer than we know.
Agencies have likely been using this and voice faking technology for years now.
The CIA doesn't want the pubic to know how long advanced AI technology has been around. They also don't want you to know the huge amount of bots out there influencing public opinion (also with AI). It's a huge psyop.
When a tech is “just released”
That means its been in use behind the scenes for far longer than we know.
Agencies have likely been using this and voice faking technology for years now.
The CIA doesn't want the pubic to know how long advanced AI technology has been around. They also don't want you to know the huge amount of bots out there influencing public opinion (also with AI). It's a huge psyop.
This media is not supported in your browser
VIEW IN TELEGRAM
Chinese IT giant Baidu introduces its own ChatGPT alternative.
Baidu's founder promises that their bot can perform up to 40 billion operations per second, generate videos and images, understand user behavior, and simulate the logic of human thought and responses.
This creation is set to be released in March, according to statements, and the market capitalization of the project has already reached $3 billion. Whether it will surpass the current version or not, we don't know, but it has certainly stirred up some buzz in the country.
Will the Chinese surpass GPT-4?
Please open Telegram to view this post
VIEW IN TELEGRAM
Now that AI generated images and other content is mainstream, we can safely assume that those strange videos where Biden never looked quite right, are in reality a more advanced version of AI tools
If you don't know already, we now have access to powerful AI tools to create images that look perfectly real.
But as powerful as they are, the government and rich people have access to even more amazing versions of this.
They have used these to create near perfect videos that billions of people watched. But there is a catch. For a short period it was not 100% perfect. Remember when Biden never looked quite right?
Always a bit different, his ears where different from video to video, maybe some strange flicker in the eyes? Or how he always was in the studio instead of the traditional White House location?
This applies to Putin too. The russians have a similar technology. Yet not quite right. At that time. but now? I think now they have perfect tech.
If you don't know already, we now have access to powerful AI tools to create images that look perfectly real.
But as powerful as they are, the government and rich people have access to even more amazing versions of this.
They have used these to create near perfect videos that billions of people watched. But there is a catch. For a short period it was not 100% perfect. Remember when Biden never looked quite right?
Always a bit different, his ears where different from video to video, maybe some strange flicker in the eyes? Or how he always was in the studio instead of the traditional White House location?
This applies to Putin too. The russians have a similar technology. Yet not quite right. At that time. but now? I think now they have perfect tech.
This is how I think AGI will hit.
I think AGI will be achieved by some average guy at home with excessive money to buy equipment, using open source models, enough programming knowledge to alter/free code, and either a lack of foresight or malicious intent. An LLM (or next gen) trained on black hat coding and given unfettered access to the internet could do more damage to world infrastructure in seconds than all the malicious hackers in the world combined. Especially if you let it learn and recode itself.
I think AGI will be achieved by some average guy at home with excessive money to buy equipment, using open source models, enough programming knowledge to alter/free code, and either a lack of foresight or malicious intent. An LLM (or next gen) trained on black hat coding and given unfettered access to the internet could do more damage to world infrastructure in seconds than all the malicious hackers in the world combined. Especially if you let it learn and recode itself.
- OPEN AI CEO won't give away ChaGPT's secret recipe: "This could get super-dangerous very quickly"
- Google's Med-PaLM 2 medical LLM passes medical exam with flying colors
- "Stanford's Alpaca" shows OpenAI might have a budget problem, its AI would cost too much.
- U.S. Copyright Office: AI-generated art can now be copyrighted.
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
Sam Altman admitted that the capabilities of ChatGPT are actually greater than they appear. In his opinion, there might be people who will remove all restrictions from neural networks. And then the consequences could be completely unpredictable. He believes that this technology is changing society and can accumulate potential very quickly.
Altman thinks that they should anticipate possible consequences and prepare for them while there is still time
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
"No camera required"
Not even video input needed.
Write down what's on your mind to create animations in 5 seconds.
(Gen-2 from runwayml)
Not even video input needed.
Write down what's on your mind to create animations in 5 seconds.
(Gen-2 from runwayml)
Five websites to detect if something has been written by an AI
→ writer.com/ai-content-detector/
→ huggingface.co/openai-detector/
→ platform.openai.com/ai-text-classifier
→ gptzero.me
→ originality.ai (paid)
→ writer.com/ai-content-detector/
→ huggingface.co/openai-detector/
→ platform.openai.com/ai-text-classifier
→ gptzero.me
→ originality.ai (paid)
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM