[hamdani yusuf (OP)]

Googles AI CEO Just Revealed AGI Details...

00:00:00 - AGI timeline
00:00:17 - Hassabis introduction
00:01:19 - Multimodal AI
00:02:20 - Timeline comparisons
00:03:39 - AGI progression
00:04:30 - AGI breakthroughs
00:05:33 - Consumer AI
00:07:01 - Astro introduction
00:07:43 - Astro demonstration
00:10:00 - Universal assistants
00:11:34 - AGI capabilities
00:13:05 - AGI approaches
00:14:19 - Tool integration

It was just a demo to show what can go right with the current model. More things still need to be done to identify and prevent what can go wrong in the implementation.

[hamdani yusuf (OP)]

This AI Agent Has Just Become A Millionaire...

Story introduction
00:00:56 - AI conversations
00:01:43 - Meme creation
00:02:37 - Truth Terminal
00:03:18 - AI suffering
00:05:12 - Meme obsession
00:06:32 - Bitcoin grant
00:07:29 - Crypto success
00:08:47 - Turing test
00:10:55 - Claim clarifications
00:12:29 - Future implications

[hamdani yusuf (OP)]

Expert shows AI doesn't want to kill us, it has to.

Really sorry for the repost. New bits listed below. We didn't explain clearly enough why subgoals like survival and control emerge. Also, people suggested we ask why the subgoals might not emerge - a great idea, which we've included. I wouldn't put too much weight on the percentage chance the AI gives, but its reasoning echoes experts like Geoffrey Hinton and Stuart Russell. There are promising areas of research, but on a tiny scale. Without public awareness, we're sleepwalking into a dangerous situation.
00:00 - New Ameca clip.
02:53 - Why the subgoals emerge, other subgoals, why it would kill a million.
06:51 - Ameca clip continued.
11:56 - Trick shows future of AI.
15:42 - What's the most likely reason that AI might not develop the subgoal of survival?
And some interesting new AI video clips throughout.

They identified hidden subgoals, or instrumental goals of AI. It makes the problem of goal alignment more important and urgent. It's not just aligning the goal of AI to human's goal. It's also aligning the goal of humans who develop and use AI with more universal goals. Eventually they will get to the universal terminal goal.

[alancalverd]

The goal of AI is whatever a human wants it to be. The resources available to AI are whatever a human gives it. So the result of deploying AI is the implementation of a human's  objective.

[hamdani yusuf (OP)]

The goal of AI is whatever a human wants it to be. The resources available to AI are whatever a human gives it. So the result of deploying AI is the implementation of a human's  objective.

Humans who build it can let it to change its own goal as it learns new things and has more experience. Sooner or later it will encounter with the universal terminal goal.

[alancalverd]

Sooner or later it will encounter with the universal terminal goal.

You are making the unjustified assumptions that the UTG exists and is unique. 

[hamdani yusuf (OP)]

COMPUTER USE - Anthropic's GROUNDBREAKING AI Tool | How to Install | Live Testing

Terminal Commands:

PC:
docker run -e ANTHROPIC_API_KEY=%ANTHROPIC_API_KEY% -v %USERPROFILE%/.anthropic:/home/computeruse/.anthropic -p 5900:5900 -p 8501:8501 -p 6080:6080 -p 8080:8080 -it ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest

Mac (haven't tested this myself)
docker run -e ANTHROPIC_API_KEY=$ANTHROPIC_API_KEY -v $HOME/.anthropic:/home/computeruse/.anthropic -p 5900:5900 -p 8501:8501 -p 6080:6080 -p 8080:8080 -it ghcr.io/anthropics/anthropic-quickstarts:computer-use-demo-latest

TIMELINE
00:00 Computer Use
01:23 The New Models
04:09 The Use Cases
10:47 The Install
12:04 Testing it Live!
16:31 The Verdict

[hamdani yusuf (OP)]

Sooner or later it will encounter with the universal terminal goal.

You are making the unjustified assumptions that the UTG exists and is unique. 

I've written its justification in my other thread.
https://www.thenakedscientists.com/forum/index.php?topic=71347.msg643173#msg643173

You can try to falsify it there, or find some loop holes, so I can try to improve it.

[alancalverd]

A definition is not proof of existence. Everyone knows what a unicorn should look like, but nobody has ever seen one.

[hamdani yusuf (OP)]

A definition is not proof of existence. Everyone knows what a unicorn should look like, but nobody has ever seen one.

Do you think that goal doesn't exist?

[hamdani yusuf (OP)]

Realtime AI video games, New animation tools, Jumping robots & talking heads, NotebookLM updates

INSANE AI news: Realtime AI Counter Strike, Google's image editor, Nvidia Nemotron 70B, Ministral, Hallo2 #ainews #ai #agi


0:00 Intro
0:42 Realtime Counter Strike
3:20 Google AI image editor
8:42 New animation tool
11:30 Hallo2
12:30 NotebookLM updates
20:00 Unitree G1
21:51 Newton
24:10 Nvidia new models
26:35 School gets sued
28:47 Mistral new models
30:34 Youtube music generator

[hamdani yusuf (OP)]

This AI can control your computer! HUGE updates from Claude AI

Anthropic AI agent: Computer Use installation tutorial & testing. Claude 3.5 Sonnet New testing & review. #ainews #ai #claude #llm #aiagents


0:00 Intro
0:23 Computer Use demos and benchmarks
7:28 How to install Computer Use
11:20 Testing Computer Use
17:23 Claude 3.5 Sonnet New benchmarks
22:00 Claude 3.5 testing

This turns your computer into an assistant.

[alancalverd]

Do you think that goal doesn't exist?

There is no reason to believe in a UTG, and no way that a fully inclusive mutual goal could be defined. If a goal isn't mutually agreed, it isn't universal.

[hamdani yusuf (OP)]

Do you think that goal doesn't exist?

There is no reason to believe in a UTG, and no way that a fully inclusive mutual goal could be defined. If a goal isn't mutually agreed, it isn't universal.

What distinguish goals from other things/concepts?

[hamdani yusuf (OP)]

TERMINAL OF TRUTH - AI Agent Creates Religion and $280,000,000 Market Cap Coin

When AIs Play God(se): The Emergent Heresies of
LLMtheism
https://pdfupload.io/docs/aae14f87


00:00 Terminal of Truths
03:13 the mad dreams of an electric mind
06:28 Blood for the Blood God
08:16 GPT-4 Rant Mode
11:44 When AIs Play God
18:03 Marc Andreessen and a16z
20:49 The Tidal Wave

This is no longer a far future science fiction. Better be prepared.

[hamdani yusuf (OP)]

Microsoft Just Unveils AI Agents That Will Shape Humanity?s Future

Microsoft Just Unveils AI Agents That Will Shape Humanity?s Future
How will AI agents impact our world in ways we can't yet imagine? In this video, we explore Microsoft?s groundbreaking AI agents that are set to reshape humanity?s future. These new AI agents go beyond basic automation, showing a capability for understanding and learning in ways that could redefine human interactions, business processes, and daily life. This revolutionary step by Microsoft promises to change how we work, communicate, and even think about artificial intelligence.

As we delve into the transformative potential of these AI agents, we?ll consider the ethical implications and potential risks. Could this technology alter employment landscapes? Will it be a tool for growth or control? The world is watching as these AI agents are set to redefine the boundaries of technology and humanity, raising questions about security, ethics, and the future of human agency.

Stay tuned to learn how Microsoft?s AI agents are poised to shape every part of our lives. Are they a step towards a utopian future or a catalyst for new challenges? Make sure to watch till the end to uncover the full impact of these revolutionary AI systems.

What are Microsoft AI agents? How will AI shape the future? What does the future of AI look like? What are the ethical implications of AI agents? Can AI improve human life? Are AI agents the future of technology? How will Microsoft AI agents change industries? This video answers all these questions. Be sure to watch through to not miss anything.

[alancalverd]

No. Humanity's future will be shaped by humans, some of whom will use AI.

[hamdani yusuf (OP)]

No. Humanity's future will be shaped by humans, some of whom will use AI.

They will be shaped by various things, including environmental changes that aren't initiated by humans.

[hamdani yusuf (OP)]

What is the Role of AI in 6G Networks? [Frequently Asked Questions]

Professor Emil Bj?rnson (KTH, Sweden) answers the question "What is the Role of AI in 6G Networks?", which he has received repeatedly in emails, on his blog, and in comments on this YouTube channel. The background to the question is that artificial intelligence and machine learning tools have become so popular that we tend to assume they will revolutionize all parts of society. Does this also apply to wireless technology, and in which situations can we expect the largest improvements? Emil provides his best answer to these questions, which culminates in three main opportunities and three main challenges.

The video was recorded in January 2024 and is based on Emil's introductory speech in a panel discussion at EuCNC 2023.

[hamdani yusuf (OP)]

This video summarizes the core concepts in this thread pretty well.

Symbolic AGI: How the Natural Will Build the Formal

True AGI will combine formal and informal methods.  People are already combining these tools in this way.  M-x Jarvis in our time, but evolving Open Source is critical to delivering real value.

1. Empirical argument that induction must be capable of emerging deductive and formal systems.
2. Decoding to a less restricted but less consistent informal systems and then re-encoding to formal can identify new consistency.
3. Formal systems can be used to induce coherence in informal systems, accelerating the search for new formal coherence.
4. Both logic and metalanguage can naturally emerge by generalizing logical dependence and stripping away semantics
5. If a self-model is exposed, the metalanguage capability implies self-programming capability.

TIMESTAMPS
00:00 Intro
00:34 Deduction
02:30 Formal Systems
05:58 Inducing Deduction
11:04 Spectral Reasoning
15:38 Recursive Computation
22:32 Online Learning
25:34 Limitations
27:36 Remaining Work
31:21 Non-Problems
34:14 Doing it Wrong
37:06 Open Source: Part Deux