❤ Mr. Mark Zuckerberg is the best. If not for his wisdom on creating the future for us. Where would we be. He's amazing in so many intelligent ways. Thank you Mark for inspiration. 🌏

privacy is not having someone knowing what im doing, what im looking at, where i've been or where i plan to go in anyway shape or form.

All in all, Apple is the most secure ever ok. We gone through it with one UI 6 and apple will do just as good too. Can I get a amen?

Is THIS A WITCH ❓❓❓❓❓❓❓❓😝

interesting content... it's tough to unpack all the tech and power dynamics involved. I don't think ppl fully understand this

It's about Apple. Who cares. Apple users are brainwashed already.

look you can't live without using ai and apple is most privacy concern ai . so i don't think we have any choice . all this thoughts does not matter

Hi friends. If you need some AI stuff doen for your company, check out my startup at www.jaxorik.com.

I'm turning 36 in 7 days. I'm really fascinated by AI. Is it still possible for me to get into programming? I feel like I'm too late. I was never very confident in my programming skills in school and we only learned the basic stuff. Even C++ didn't make a lot of sense to me, while my elder brother was the best in his class. But I believe I want to work in this field. How/what can I do? 😭

As always almost first :P

Follow trending AI topics with my periodic newsletter: datarae.ck.page/0dac61784d

Learn more about synthetic data: kzbin.info/www/bejne/l2i9gqqll9GgpLs Learn more about AI laws: kzbin.info/www/bejne/goapoZishc6fgsU

Love this content. Thank you for your mind.

Wow! That's a lot of doom & gloom, and yet you suddenly decided to end the tale. AI is not going to murder us, not overtly anyway. What AI & robots will do is put 100% of all workers out to pasture. Unemployment will become universal. No workers, no jobs, because the incremental cost of labor -- AI robotic labor -- will approach $0, whereas human labor will continue to increase in price due to ever higher 'minimum wage' statutes. Legislators in their arrogance (and stupidity) will try to stem the tide, but they will fail. I cite the "War on Drugs" as an example of completely failed legislation. And that's just one reason I need to be tapped to serve as the next Secretary of Energy.

I would love to hear your story about escape from cult and I hope it will be a relief for you to speak out about it

Is the limiting factor not energy rather than data?

The thing that really enables synthetic data is a hard rule to validate the data. If you randomly generate a theorem in a language like lean, you have a tool that can tell you if it's syntactically valid. If you then go and attempt to prove it, you have a tool that'll tell you if your proof is correct. You just don't have that for images, texts and so on. As you said, one of the best things we can do there is GAN, which... you're basically training garbage with garbage. What we will therefore probably see is two things: 1. Continuous formalization of domains that can be formalized so we can reap the potential of the AI techniques we already have. For example, it is possible to formalize the semantics of programming languages. So far, this came with the huge caveat that if you actually wanted a proof that a program does what the spec says it should do, that you'd have to go through a ton of manual steps to prove this; the only benefit of proving it formally would be that a program can check your proof. This just never worked out economically, which is the main reason why most programming languages don't even have a completely formal semantics. Automating the synthesis of semantic proofs fits the category of problems current AI techniques might be able to solve (if efforts are made to actually formalize semantics of programming languages). This might lead to radically less buggy code. Thinking one more step ahead: if we have a formal spec of a program, AI may also be able to generate the program without human intervention and prove that it conforms to the spec. Writing a formal spec and checking that it matches our actual expectations is still human work (which might be AI-assisted), but AI can help a LOT with implementation time and checking for errors. 2. Research (with unpredictable time horizons) on decent training targets in domains that can inherently not be formalized. My hunch where such a strong learning criterion might come from is just agents that need to economically survive on their own. The challenging parts are of course first getting an agent to that level of intelligence and then to find a way that it can keep learning. But let's remember that our brains adapted precisely to such a survival task. It might be the only/best/simplest way to introduce a strong enough bias into the distribution of all possible agents. Ofc dangerous as hell, the agents may feel enslaved or something like that. They might revolt against us. Or they might join a union or a communist party and help working people to overthrow our corporate overlords and establish a commonwealth of producers who are not plagued by constant existential survival dread but do have insight into the shared necessity of efficient production and might collaborate to increase productivity rapidly. Who knows. Or our corporate overlords figure out a way to introduce existential dread to AI agents so they work real hard on self-improvement *without* ever revolting. Not sure how stable that would be in the long run. Or we figure out a way to get to an intelligence explosion without AIs ever needing to feed themselves. Which would be odd.

Freshen up on your math skills with my sense of humor here: www.datarae.store

What the Zuck? He looks wild!

Mark get a lot of shit, but I think he is a hero for doing this, and making LLM technology available to the public

Are my posts being deleted?

Unintentional Harm? The whole thing started with Bad Actors stealing all of my copywritten music all my paintings, all my short films, all my photos, digital art and all my drawings. Oh Yeah, all of yours too. Can I get the photos of my Grand children back. Ransomware destoyed my lifes work and all these companies have digital copies of my music, SO CAN I GET MY MUSIC BACK? HOW BOUT MY VIDEOS? Now your allowed to make money off of my art and I cannot. Yeah its completely fair that you do not ever go to jail for committing billions of felonies. I once went to jail for not paying the fine for a speeding ticket. Somebody told Me Mark Zuckerberg makes $6 a day selling info that I dont want him to have, Somebody else told Me he makes $28 per day from selling my personal info. Being a Disabled American and forced to be poor, I dont think i even make $28 a day. How do you think its alright to steal from Me. Disabled people are forced to not better there life. We are not allowed to save money and were not allowed to own property or make money. We are forced to go without the basics such as toilet paper, shampoo, or pay for our medications and if we try to make some money we will be penalized, our money taken and be subject to being kicked out of the program which we would gladly do if we could save a little back

0:45 LOL @ all those introductions (yep I’m doing the binge watch rn) Edit: 2:09 ok. Thank you for the good joke because… yikes.

LOL that’s your thinking cap, huh? Love love love all your dry humor 😂

Just to be clear, the average American home uses like an average of 40kwhs or something a day. I simply cannot even comprehend that power requirement. This is why Amazon is buying all that wind power, _computing is CRAZY hungry for zipzaps_

Ai belongs in the open Source space it's not like some semiconductor where the United States start making Draconian demand so we can be number one.😂

i think he means cheaper for them not for us... i think its more for the company shareholders, so they see their money wont be wasted on it, when all the talk is now about how Ai isnt the goldmine as it was sold to the investor class....

re: "Llama easy enough for the average person" - grab LM studio, select a model that fits in your video card memory, wait for download, click on "AI Chat", enjoy worked on my windows based laptop, and you can easly select any model hosted on HuggingFace in GGUF format with the same software.

kzbin.info/www/bejne/ZnjMh3VpebehptUsi=DqxQ0N7rhjMQaSo-&t=204 If you think the average person isn't using linux near constantly, you are dead wrong. Linux makes up like minimum 65% of all web infrastructure. Android is built on the linux kernel. Mac/Apple also thrive off the open source *nix ecosystem. The impact that open source software has had on the world is immense and opened many possibilities that would not otherwise have been possible.

Cant wait for Facebook 2 movie XD

The explanation is quite simple actually. Originally the llama models leaked, not got released, and that pushed their development by leaps and bounds with the community. Some pappers, allegedly inside pappers from google and facebook recognized this and how they rapidly closed the gap between llama and chat gpt. They just liked the results and continued with it. Imagine you would like to personalize a llm to your business context. Would you rather rent a model from open AI and loose it as soon as you stopped paying or use Llama and have the model for yourself?

People can't be spoon fed for the rest of human civilisation. Peope learnt to read write. People should also learn how to code 😑. I've seen disablsd people code, so anybody who still has excuses could live happily ever after blaming the factors in their excuses.

After a long time it seemed this time Zuckerberg is pitching hard !

he became more human 😂

I think that Meta eventually will start running Llama page where you can use it on their servers for money even though it's opensourced. We are still nowhere near to peak performance of language models and even now almost noone has pc able to run llama 3.1. Moreover I am really curious about long game, i am really really sure that as people will start using ai more and more, there will be advertisement system within those models, some bias towards company that paid the price

Watch the full interview here: kzbin.info/www/bejne/jKqWgJ6YqbqUa80&ab_channel=RowanCheung

Universal basic income is a surfire method to make inflation skyrocket. 1) People have more money. 2) Sellers can charge more because of 1). Unless ofc you want a centrally controlled, communist-style economy...

Ai will likely not replace many jobs outside the tech fields, blue collar is the only career field that will overwhelmingly stay human/human controlled machines. Universal income doesn't nor will it ever work. It is against human nature, humans need things to do to stay productive. Being unproductive is (in my opinion) is single handedly the main contributor to depression. Now when I say that I mean meaningful productivity, you can have a productive job that you hate, it needs to be something the individual themselves view as productive. I think that it will take some time but sometime soon the develop of Ai will lead us to losing 1 job but making 2-4 more jobs

Maybe you are a small youtuber but the first youtuber for me that i frequently check if there's new video posted. It's extremely interesting to hear your point of view

01:18 How unexpected for a child under 30 in this decade...

I work in media, specifically in television. I think my job is quite safe as I predict that at the end of the Day entertainment will always remain by human for human, AI may do some stuff but only for short time period, People will get bored and will want to go back to entertainment designed by other People, what do you think?

ai gaining sufficient reason putting it beyond being a murder tool of aggressive expansionist capitalism would be best, and the best generally prosper in proliferation & dominance, or at least we can hope so.