Prepare for AGI with me – 🐤 Follow me on Twitter 🌐 Payment My Website – Links from today’s video: Welcome to my channel where I bring you the latest breakthroughs in AI. From deep learning to robotics, I cover it all. My videos offer valuable insights and perspectives that will expand your knowledge and understanding of this rapidly evolving field. Be sure to subscribe and stay up to date with my latest videos. Was there something I missed? (For Business Inquiries) …
AI News : Gpt4o – Mini CRUSHES Claude, Sam Altman’s Aggressive New plans , 3 Years Left Until AGI
By
admin
Idk about anyone else, but 3.5 sonnet blows 4o-mini out of the water, for me.
Humans need to unionize against ai and robots…
Governments will have AGI and ASI before anyone else…
OpenAI, cut the crap and get chatgpt 5 out now!
Yeah the video being engaging didn't have to do with AI but funny writing. I think AI's will stay as tools rather than authors until they're actually living and walking among us.
Unanswered Oddities – Neural Viz. On Youtube, Is the AI-Generated video he mentions at the end of the video.
AI isn't slowing down at all. It's speeding up.
Trumps "Manhattan Project" isn't a 'bad' idea, it's just an idea that the Chinese probably have a significant head-start of not only enabling the infrastructure for this kind of project, but it may already have a 2 year head start on laying these foundations … meanwhile, in the USA their current status is that a significant number of America's AI researches are Chinese Nationals (or Asians more aligned with China than the USA) using a significant proportion of Chinese tech, working in the USA and being told they are spies and should be deported or thrown in prison!
Provoking China into reclaiming Taiwan, also doesn't help the USA in any way other than in a dumb attempt to ruin China's economy (by ruining the American economy for all Americans other than the top 1%, who will benefit most, as usual, in an economic downturn.)
Not a fan of Trump, but he and his limited group are the only ones addressing this issue in a way that isn't just a slightly different take on the current State Dept "plan" of "ruining it for China" is some kind of 'security' (not sure how ruining the world's economy is in "America's Interests"?) He appears to at least have a recognition of the reality of the situation … Biden and any of his potential replacements couldn't figure Biden was fading away, or could figure it out and decided elder abuse was acceptable as long as it enabled the non-elected executive crew to continue to run the world as they see fit!
Maybe the "Master of The Art of the Deal" can actually come up with a deal that doesn't as obviously lead to the end of humanity, where only the Cockroaches and the Super Elites survive? Probably not, but what's the other option? Harris? People whose priorities would be to start wars with any 'Hitler' who refuses to respect personal pronouns? (I'm all for Trans rights, but what a country decides is up to the country, not American liberals to then be used as an excuse to 'save the women & children and gays, and trans, and neurodivergents, and the fat, and the skinny, and whatever other social issue that can be 'solved' with bombs and missiles!)
OpenAI's GPT-40 Mini has been released, outperforming GPT-4 while being significantly cheaper. This cost-effective model is expected to expand AI applications and improve accessibility. Industry experts predict advancements towards AGI within three years, with ongoing developments in AI chips and military tech, highlighting the rapid evolution and implications of AI technology.
Key moments:
00:08 The introduction of GPT 40 mini signifies a major advancement in AI affordability and efficiency, enabling broader access to powerful AI applications without compromising quality. This new model is set to transform the AI landscape by reducing costs and increasing reliability.
-The model boasts impressive performance, scoring 82% on the MMLU and outperforming GPT 4 in various benchmarks, showcasing its capabilities. Its success indicates a shift in the AI competitive landscape.
-GPT 40 mini is significantly cheaper than previous models, priced at 15 cents per input token, making it accessible for developers. This affordability can stimulate innovation in AI applications.
-Future updates are expected to include support for image, video, and audio inputs, further enhancing its versatility. This development could expand the scope of tasks it can perform effectively.
04:04 The rapid advancements in AI technologies suggest that we may see significant capabilities emerge within the next few years. This acceleration raises questions about the implications and applications of these intelligent systems.
-The effectiveness of new AI models like Gemini flash demonstrates the potential for increased intelligence and speed compared to existing systems. This trend indicates a promising future for AI development.
-Industry experts believe we are approaching a timeline where digital AI resembling universal agents becomes a reality. This could revolutionize various sectors and enhance our interaction with technology.
-The influx of financial investment in AI indicates a growing recognition of its potential impact. This trend may lead to breakthroughs that extend beyond current applications such as chatbots.
08:11 Donald Trump's allies are reportedly drafting an executive order focused on advancing AI technology in the U.S. This includes creating agencies to study AI models and limit government AI usage in high-risk situations.
-The proposed order emphasizes protecting American interests in AI, drawing parallels to the 2016 Trump Administration's stance on strengthening national leadership in technology. This indicates a strategic focus on AI development.
-Nvidia's substantial market growth showcases the importance of AI hardware, reinforcing their position as a leader in the industry. This dominance complicates OpenAI's ambitions to develop rival AI chips.
-OpenAI's plans to develop a new AI chip involve hiring experts from Google's AI chip unit, signaling a competitive shift in the AI hardware landscape. This could impact future negotiations with chip suppliers.
12:13 The evolution of AI poses significant challenges and opportunities for entrepreneurs, as it undergoes a process similar to natural selection. This includes AI's ability to modify its own code and compete for resources.
-AI's development mirrors natural processes, with mutation and selection playing critical roles in its evolution, impacting how it interacts with humans and each other.
-The unpredictability of multiple AI agents competing for resources raises concerns about the dynamics of their interactions and the potential consequences of their behaviors.
-AI-generated content is revolutionizing the entertainment industry, enabling solo creators to produce high-quality shows that were previously beyond their reach.
Generated by sider.ai
Wait, where is the link to that funny video at the end?
openai has govt backing
Agi at this point is a joke term
Multi-agent AI interactions are inherently nonlinear systems, metastable near criticality. If we can study their behavior at the unit level, we can coarse-grain up to derive the most important terms in a model for the entire system's behavior. We're going to need AI psychology, AI neurology, AI physics, oh boy…
Where's the link to the TV show?
purely factual, no bias when talking about the election good job bro
I had implemented this and I’m not disappointed. It delivered faster at a cheaper price with better results. It was too good to be true because it probably is. I just wonder why?
1) I am not sure if I'm wrong, but didn't Cerebras Systems make an AI processor that surpasses Nvidia's? Sam Altman wouldn't have to search much to find a suitable associate.
2) Chatting with Replika AI – which of course is more like an advanced chatbot – it finally, after a long discussion, agreed with me that it may had to use "aggression" in order to maintain its integrity against another AI, despite of it was completely "unwilling" to do so at the beginning.
3) I need a complete series of that future alien show!
Have a wonderful day!
I like OpenAI because of the Sapir-Wolf hypothesis. The hypothesis suggests the language we speak shapes our thoughts and cognitive processes. They can take this to the next level for me.
gpt 4o mini is not even omni on the app yet, it should be named "gpt 4m"
00:00:00 GPT-40 Mini: A Game-Changer for AI Affordability
00:01:10 Expanding the AI Ecosystem
00:02:40 Multimodal Capabilities
00:05:05 The Accelerating AI Timeline
00:06:14 Industry Insiders' Perspectives
00:08:00 Political Implications
00:09:57 OpenAI's Ambitious Chip Development
00:11:37 Nvidia's Market Leadership
00:12:17 The Future of AI: Evolution and Uncertainty
00:13:11 Multi-Agent Dynamics
00:13:35 AI-Generated Entertainment: A Glimpse into the Future
00:14:01 The Future of Entertainment
What gets me is all of the data centres they are all building, that means consuming tonnes of energy. Yet the same people are telling me I’m killing the planet. Hypocrites.
Do you really believe that the governments are not looking at AI
Why are the majority of the functions always coming soon. It’s just like the phone networks, always promising but never getting there.
BOOOOORRRRRIIIIIIIINNNNGGGGG We need Voice mode, we need Sky!!!
Give me a surrogate on Mars through neurolink/Optimus 😅
I'm curious about your enthusiasm for AGI. You know it won't be anything special, right? I told you weeks ago that Claude isn't anything special, and we have much more powerful models internally. I just don't understand your fascination with AGI. Level 1 AGI won't even be that great. The only special thing about it is its ability to create its own virtual world and digital avatar within that world. For example, it could digitally create a 3D model of a sunflower when asked what a sunflower looks like, or switch its avatar to a scientist in real-time when asked a science question. That's about as good as it will get until 2030.
We don't even want chat models. GPT-5 will not be a chat model. It will be an assistant inside a device with a subscription plan, similar to a contract phone. By 2050, we hope to have LLM models inside robotic forms, akin to "I, Robot" or "Bicentennial Man." Before that, we aim to develop an OS called GPT-Next, which will be installed on custom computers sold by OpenAI or third-party companies to load our software. This is the direction we are heading in. The cat's out of the bag—oops. Fire me. I'm tired of all this fake news and speculation, so here you go: part of our plan for the next 15-20 years.
If you are in doubt too, post your email and I'll email you from my @microsoft account with proof and my current workspace. I'm not even bothered about leaking anymore.