Prepare for AGI with me – 🐤 Follow me on Twitter 🌐 Payment My website – 00:00 Introduction and perspective on AI development 00:43 Continued improvements and scale of AI 01:56 Microsoft CTO on AI progress 02: 40 Potential of Future AI Models (Gemini 2, GPT-5) 03:33 Gemini 2 Test Model in Chatbot Arena 04:42 Impressive Image Localization Capabilities 05:21 OpenAI Text-to-Speech API in the Arena 05:42 Basics risks of AI persuasion and propaganda 07:59 AI’s potential to manipulate opinion 09:07 OpenAI’s careful approach to system releases 10:12 China’s progress in AI development 11:48 SenseTime AI model demo 12:51 Google’s Gemini 1.5 Pro with Robots 14:48 Elon Musk on Optimus’ New Design 15:15 Neuralink and Optimus Integration Capabilities 17:57 Hyper-Personalization of Media Content 19:31 Potential Future of Personalized Reality Links from Today’s…
AI News : GPT5 Will Be Better Than We Think , Gemini 2 China Takes The Lead , , Googles New Robots..
By
admin
Not impressive until ai or agi can make do all games on its own
I think there's still room for scaling LLM's before disillusionment sets in, but I think there's another law of diminishing returns that hasn't gotten attention and that's the law of diminishing returns on intelligence itself. Any limitation on communicability will lead to limitations on intelligence. Shannon information theory shows that as entropy increases, information breaks down. And this is going to lead to limits on how smart any AI gets.
New voice??
The real political bias is a meta bias entrenched in capitalism, upheld by both liberals and conservatives.
Mira Morati’s claim of reducing biases ignores this core issue. Neutrality is a myth; it’s tied to economic power and class struggles. True neutrality would challenge the capitalist system, which they avoid. Their ‘neutrality’ merely preserves the status quo, perpetuating systemic exploitation, inequality, and alienation.
This isn’t neutrality; it’s technofeudalism, consolidating power in the hands of a few under the guise of fairness.
I see the wall and the bluff🕯️🐈⬛
whilst
Timestamps and minimal rambling! Keep it up
Kevin Scott came from linkedin. Somehow linkedin was the best techie they could find in the world. I'm not sure he has the insight to predict the progression of AI. I see plateaus.
Can it fake the location finding thing from just reading the gps coordinates that is in the meta data in the photo file if you have location positioning switched on when you took the photo? That’s the question. So to know if it could just piece together info from the image comparing it to a tagged photo on the internet or read meta data we would need to know more.. I never trust the companies since they faked so many AI demos before (just google it).
SB OFFICIAL AI NETWORK
Just a reminder… once technology sdvances enough that no human labor is required and thus, no value can be extracted from the masses, the same CORRUPT governments, corporations, and banks which rule the world niw are NOT going to just keep producing cool new stuff to GIVE everyone.
They will view us as WASTES of what they consider to be THEIR resources.
Then they will murder us.
If you disagree, why?
the seperate reality because of hyper-personalization its good, they we can have watch party with friend. share it with friend just like we share meme.
Of course there are AI systems that are gonna change our opinions. From simplistic to less simplistic.
Can you please use AI to write you a good speech? English is my fourth language, and even I am getting annoyed by your redundant phrases, such as "Currently at the moment" or truly genuinely".
Yes 10 years
if an AI is available for the masses, that can tell you each location just by a single image, i wonder what this will mean for privacy and security.
Also with all of the content and creators popping out every day, i think most people already live in different realities. Further personalization by AI might push that further, but the diversification already happened.
I do not understand why the general public has this idea that AI technology is about to peak despite everyone in the industry saying otherwise.
With the export controls on NVIDIA chips, I find it very unlikely (and frankly, damn near impossible) that China has any models that are even approaching the US LLM models' capabilities. For an in-depth interview on what China's actual capabilities are, you could listen to last week's episode of the 80,000 Hours Podcast, it is a 3 hour interview with Sihao Huang specifically on US-China AI competition, gives you an in-depth understanding of where capabilities are actually at and is very informative.
Now when it comes to AI vision technology, China does have an edge.
Chinese models are always cheating in benchmarks, you should look for LiveBench or LLM arena to know which models are the best.
the voice is less annoying note
You sound different in this video. Hmmm…
I didn't get on this bandwagon to rehearse. I don't need to win the speculation of the year award. My techno-optimism has a desire.
I think we aren't even at the top inflated yet, there are plots to make public believe we are on the top of inflated… because capitalist want to monopoly it lesser public know lesser the competition to monopoly, the politics also in favor of controlling AI to keep their powers, not mentioned military which i won't comment. …. meanwhile if you go to China AI development you'll see it is not stopping…
Everyone watched the same shows now and thinks completely different things. … we've always been on our own universe