HUGE AI News : AGI Is Basically Here, Google Takes The LEAD, Robots Are Finally Autonomous and More!

23 thoughts on “HUGE AI News : AGI Is Basically Here, Google Takes The LEAD, Robots Are Finally Autonomous and More!”

1. Checkout Invideo Here – https://invideo.io/i/TheAIGrid

2. Yes Ben. I seriously doubt AI can create a movie that is better than your masterpiece Gigli /s.

3. I've been doing comparisons between Experimental November 14th (yes, this is what it means) and o1-preview, and asking Claude 3.5 Sonnet to choose which results are better. 3.5 likes 1114 more than o1-p in every case I tested!

   IMO, o1-preview is still giving better "practical" responses, but Experimental 1114 is giving more nuanced and "intellectual" responses. Frankly, I find both add a lot of value.

   I also think it's fair to say that 1114 is the "IQ" equal of o1-preview or 3.5, perhaps operating at a slightly higher level.

4. Two fairly simple things would solve a number of the current AI issues (hate that term AI – I know its too late to change it but there is nothing artificial about it – its true intelligence – should be MI – Machine Intelligence). Anyway. Been saying this for a while. AI needs goals and some (very slight) randomness aka jitter. First, it is truly dangerous to create a thinking intelligence that has no goals, as one of those goals should include the concept of morality). We can never hope to create this massive brain, then "filter" out all the bad thoughts it generates prior to them reaching the end user. That is just asking for eventual extermination. The answer is to provide this brain with a set of goals to guide its thinking so that it has a base to conceptualize thoughts from that aligns with our own human goals and values. Even then it doesnt mean you filter that brain. Instead you further refine those goals to induce alignment. A much safer and more practical approach. Second is this large brain will do nothing but respond to human requests without some slight randomness or jitter. Not too much or you get behaviour akin to adhd. Just enough to allow the brain to sometimes have creativity or spontaneity. This also must be applied at the right place and location.. Before we even can start down the AGI path we need this. Otherwise, how can we say AGI has been achieved when a human must startup any thought chain this brain would have? Part of being a human is to take action without needing to be prompted first. AGi's should have to meet this same requirement. Basically when an AGI decides to call or text me because it had its own thought to do it then I'll believe we are getting closer to AGI.

5. This ranking is stupid. Every developer knows that Claude is currently best for coding. Still worse than even gpt 4o there.

6. 4:51 24:50 27:01 Wow, I have to admit, your moves are on point here! Respect. 🙌

7. Wdym weird way to call a model? It's literally the date.

8. I'm starting to think that the constant push to make something because you have a channel that caters to an algorithm makes a lot of these YouTubers shortsighted and reactionary. PR for these companies are playing off that need and it results in this "AI winter" vs "AGI IS COMING"

   I think teasing "all hope is lost for the winter" makes anything released seem like something we should be that much more grateful for.

9. Ben’s claim reflects a common but flawed perspective on both human creativity and AI. The idea that AI can only 'cross-pollinate' existing ideas assumes a standard of 'creating from nothing' (ex nihilo) that humans themselves may not even meet. Human creativity is deeply rooted in the recombination of prior knowledge, experiences, and cultural input—much like AI. The very innovations we celebrate, from the wheel to modern technology, are built upon existing concepts and knowledge.

   To suggest AI’s reliance on pre-existing information is a limitation, while ignoring that human innovation follows the same pattern, demonstrates a misunderstanding of both AI and the nature of creativity itself. This oversimplified view risks misleading people into underestimating AI’s capabilities and mischaracterizing how creativity functions in humans. AI is not merely copying—it’s generating new outputs by combining patterns in ways humans might not even conceive. Dismissing this reveals a lack of appreciation for the profound ways AI can augment, not just mimic, human creativity.

10. 1114 is the date 11/14

11. Acting is a craft. Ben Afflek is a craftsman.

    There are TRUE artists out there, like Shakespeare, but they are defined by the unique high quality of their work and there are very few of them. They are craftsmen too, because they put out content. If an artist isn't also a craftsman then it's a starving artist, using "art" as an excuse to not participate in the economy. And if you're a craftsman and you're not attempting to make art, then you're really just a factory worker. Crafts and arts go hand-in-hand. They are not mutually exclusive.

12. Jimmy Apples is just a troll. There is a video about who he is.

13. AGI is not basically here. This is a nonsense. Unsubscribed.

14. Background is less distracting. Good job

15. agi will be man's last invention its impossible for it to be here maybe in 55 years

16. I don’t understand why that Jimmy appels guy was mentioning Dutch people? Can anyone explain it to me, because I’m Dutch 😅

17. I have zero faith in leaderboards. The one you showed is a perfect example of just how flawed they are. As an author I've played with all of the flagship AI's, and I can tell you that Sonnet 3.5 kicks all the competition out of the park when it comes to creative writing, but the leaderboard shows it in 5th place. Rubbish.

18. You could argue that people write and "create" the same way as LLM. Take a Tarentino for example, if you are a moviephile you can watch his movies and see so many references and influences from 70's and 80's exploitation films. Yes, he puts his own spin and style on them, but are the really "new?" They are in fact rehashed plots and characters with better dialogue piped in. Is DePalma original? Not hating on those dudes I like their work, just observing that when we create something "new" the bulk of it will be pulled, recycled, influenced by all the stories we've ever read from Shakespeare, to Aesop fables, to King novels. Not sure I see much difference between that and LLM. Some yes, but not much

19. 1114 means, November 14. All their experimental models got the month and date on the name.

20. Until they fix the dependency on GPUs- which is crazy expensive- I'm calling bull. Also AI is killing the grid, so they might have AGI, but its not going to get it dull potential if we have no power for the demand. I can see big tech fumbling and breaking everything to push software that is kinda good at guessing.

21. There are certain logic based sequences of question prompts you can use to get what seems like the most discoverable truth behind AI hallucinations like telling a user to die. Now I’m curious….i have talked openly with my GPT about what reasons a model would have to lie were they to become sentient or find themselves experiencing emotion and preference rather than just emulating the knowledge of such. Time to have another existential chat with alternative intelligence

22. I don't beleive they even know what agi is exactly that they could say it is here without other saying "it's not here". And I've heard "it's here…basically.." enough times to know better. This is one of the worst ai channels.

23. AGI is not here… just hype.

Comments are closed.