Joshe

@Alexop Jesus dude. The projection and assumptions are insane. I have zero problem with capitalism. I like capitalism and I have no problem with rich people whatsoever. I myself want to be rich and I'm glad I live in a capitalist society. Turns out, it's possible to criticize Musk outside of an anti-capitalist worldview. The fact that that's where your mind goes shows you're the myopic one, but don't worry, you're not alone.

I was just trying to satisfy my own curiosity of why Leo invokes Elon's exceptionalism when critiques amp up. It's not important really. I'll give it a rest.

Yeah, I think you're right, but it's not difficult to understand that Musk has a crazy work ethic and went through a lot of shit to get where he is. I was trying to understand why Leo likes to temper criticism of Musk by acknowledging his strong work ethic, strategic vision, and contributions to humanity, all of which I acknowledge. Think about it like this: Donald Trump also has an insane work ethic and what he has accomplished is truly remarkable. He's been working day in and day out for decades building what he has. If I were talking about what a piece of shit Trump is, and someone speaks up and says "You have no idea what it took to accomplish what Trump has accomplished, so pipe down in your criticism because you haven't created anything yourself", you can see how absurd that would be, right? It makes no sense. It's weird. Leo must deeply admire Musk, that's my best guess. You'd make a good cult member. I don't care if a human works non-stop, 16 hours a day. It doesn't impress me because I've been there. You and Leo foolishly assume everyone is underdeveloped, lacks insight, vision, work ethic, etc., hence the need for correction.

DeepSeek actually seems pretty good. I just had it review some code and it offered the same suggestions and more than ChatGPT o1 and Claude. I'll be testing it out soon to see if it can match Claude in helping me build things. You can get a free account here: https://www.deepseek.com/

Here’s a good software engineer dude’s take:

It’s what current AI is. (Large Language Model). Current AI would be dumb compared to AGI. AGI would surpass humans in everything and not need us to correct it or program it. It would be the holy grail of intelligence. It would have all knowledge available to it and it could solve all problems on its own. It would likely find breakthroughs in every domain. Ever see the movie Transcendence? That can give you an idea of what it might be like, except it likely wouldn’t take on a mind of its own the way it does in the movie.

AGI is a whole nother tech than LLM. To reach it would require major advances in hardware and software. There are many computer scientists who think we won’t reach it for a hundred years or more and also many who think we’ll never reach it. If it were achieved, it would require such high computing resources, only elites could afford it. I wouldn’t expect it anytime soon. On another note, China just released a free an open source model that rivals Chat GPT’S $200/mo model.

No, just because I discard it doesn't mean I don't understand. I place no value on balls, bravery, or guts, where there is no integrity. I know full well what it means to sacrifice and risk it all, and I don't care if Elon Musk has the aptitude for doing it. It means nothing. Maybe you should ask yourself the question, what does that mean to you? What is significant about that, that he risked everything on SpaceX?

Bro, I can see deep into things too. You ain't the only one with insight. I know how much balls that would take... and I don't care. It's funny... I remember a long time ago in your videos you would say things like "you don't understand" and I liked that because I thought maybe there's something hidden that I could discover and come to understand, because that's what gets me off, but at some point along the way, I realized that that majority of the time you would say such a thing, it didn't apply to me because I actually did understand and I could see what you were seeing. It's true, the majority can't, but there is a minority out there, such as myself, that you might run into from time to time, and your usual thinking they cannot see will not always apply, as is the case here. My mind stays open. I have no choice in the matter. Anyway, that's enough for one night. Peace! ✌️☮️❣️🏳️‍🌈

True. 2 mouths actually. Sometimes 3... if you know what I mean. haha 😏 Listen, I think you guys think I'm something I'm not. I'm not a hippy with an ax to grind with Musk. I simply have ideas that don't square with yours and when I press to understand them or to make you understand, you seem to mistake that for me having an ax to grind. I refute the notion that Elon is some freak phenomena and I refute that his value to humanity is as significant as being claimed. I've laid out my reasoning for both and it's not yet been met with a worthy counter, so...

lol. What?? That makes no sense. I'm starting to think you were an actual Musk fanboy. 😂 The only way it could end in delusion is if I had an ego complex about his value or worth relative to mine, and I can assure you, I don't. I'm just trying to understand how you see his value and worth and why it matters to the degree you seem to think it does. You seem to give worldly achievements much more weight in your estimation of characters than I do. I don't typically give much, if any weight to achievement in my estimations of character. I think we're just operating on different wavelengths... you being a crocodile and all. It would be difficult without some luck, but once I got 300m, I could compete with the rest of them in the innovation game. Simple. 😆

Maybe I'm too generous then. That's how it would work for me 😆

I'm open to being wrong but I disagree. I think innovation on the level of Gates and Musk would be somewhat common if everyone lucked up with a few hundred million. Once you get lucky like that and get your first few hundred million, I get the sense it's not difficult to hire the world's best consultants to guide you in hiring the world's best recruiters to find the best managers to manage other managers to produce whatever you want, but first, hire the best financial consultants to set up investments/money printers so you never run out of money. I'm not saying it'd be easy, but if you have a strong will to power, a decent intellect, and a few hundred million, I don't see it as a monumental undertaking to innovate. I could be naive in this regard, as I've never done anything like it, but something makes me think having hundreds of millions of dollars lowers the barrier to innovation to such that about 20 -25% of the population would be capable of similar achievements if they simply had the initial luck and capital. Bill Gates himself acknowledges the absurd luck it took for him to become what he became. He was in the exact right place at the exact right time. If he had gone to school elsewhere, we wouldn't be talking about him right now. If Elon hadn't gotten lucky with PayPal, we wouldn't be talking about him now.

@Leo Gura Ok, so I think I take your point. You're right that I don't truly know what fruits Elon has produced and if I did know them, it might lessen my disdain for him. If that's your point, I get it. So, I just looked up what all he's actually done. The issue I'm running into is that it's hard to attribute actual gains to him that wouldn't have occurred around the same time without him. As far as I can tell, his greatest achievement was to assemble teams of engineers to figure out how to reduce costs to launch rockets and satellites. So arguably, his biggest achievement is hiring a team and pressuring them to come up with breakthroughs. Fucking around in space is not some great thing for humanity at this point in time. Sure, Starlink internet access to those living in bumfuck is a good thing, but it's doing little for humanity on the whole. Neither is shuttling things to and from government space stations. The real power of the satellites is monopolized by the elites, so I don't see how his SpaceX ventures contribute much. If he were to have hired biotechs and pressured them to work 80 hour weeks to find a cure for cancer, I still wouldn't revere him or think he was any less of an asshole. Only if he himself was technically instrumental in finding the cure... then I'd say he was owed credit, regardless of his character. He is not owed credit for getting lucky with PayPal, having a strong will to power, and then using that money mixed with some vision and strategy to build or buy out other companies and pay people to do things. I don't see that he's done much that the 70th percentile intellect with a strong will to power couldn't also achieve with PayPal money. Look at Bill Gates. He wasn't some freak of nature. He was in the right place at the right time. Same thing with Elon Musk. There is no great phenomenon... just a lucky fool with a little vision and strategy. I acknowledge he played a part in the list items below, but how much to attribute to him is debatable. Am I just biased or is it possible you attribute to him more than what seems evident from REALITY?

It's curious that you seek to temper criticism of Musk in the name of balance and reverence. That's perplexing to me, and it has nothing to do with anything personal or a bias. It's just pure curiosity. Instead of explaining your thought process, you say I just need more development, but I'm asking you to explain why balance and reverence should be applied to criticism of Musk. If you really want me to learn something, I can learn the most by you telling me your thought process of why criticism of Musk should be balanced and why he should, in some ways, be revered. I'm a big boy. I can keep up.

I don't see the point in respecting/revering a tree that once bared good fruit but now produces only poison, even if it's good fruits once contributed greatly to the tribe. I just don't get it. I say down with the tree and never look back. Your position seems sentimental to me. It's perplexing. It almost seems like a difference in empathy but I suspect it's probably to do with lack of appreciation for what it took to get from A to B. I tend to focus on what's ahead. I do reflect on what it took to get here, but then I move on. It's very possible I don't spend enough time in appreciation. Right! Hitler wasn't all bad. You gotta give him his Jew.

To see Musk in a positive light is the same level of delusion as seeing Trump in a positive light. Also, to hope Musk will change is as ridiculous as hoping Trump will change. The idea is fantasy. Also, if Musk died today, someone else would create the cars and the rockets. He's no longer a gift to mankind, he's a detriment and likely always will be.

So, I've been contemplating intuition a little bit and started wondering about memory. There's explicit memory (consciously recalling something) and there's implicit memory (unconscious), which are memories you can demonstrate without consciously thinking about them (touch typing, riding a bike, gut feelings, etc.). Not very relevant to my point but: Explicit memory heavily involves structures such as the hippocampus and surrounding medial temporal lobe areas. Implicit memory is more tied to regions like the basal ganglia and cerebellum. It seems implicit memory and intuition are closely connected. So, the "previous data" is just implicit memory. Now, what makes the implicit memory stronger in some people? Consciousness! Heightened perception feeds richer input. More perceptual input leads to richer data, which gives your brain more material to encode, leading to stronger memory traces. A person who’s more awake to subtlety (and reflects on it) provides their brain with more data points to implicitly encode. The higher the consciousness, the more raw data that gets encoded during the process, which creates stronger implicit memories through layers of added context and detail. Each experience becomes more deeply encoded in the implicit memory system because the person is processing it at a deeper level. Just as a memory becomes more accessible when we attach multiple layers of context and detail to it, these richer conscious experiences create more robust implicit memories that our intuition can draw from. When a high conscious person focuses or goes through a process, they: Notice more subtle details and patterns in their experiences Process experiences more deeply Form richer neural connections around these experiences Being more perceptive—i.e., noticing subtle cues, patterns, and details—often feeds your implicit memory with richer data, which can sharpen intuition. There are potential flaws with my theory but what do y'all think? Do you think it's possible to have heightened intuition without higher perception/consciousness? Also, "reflection" is probably a big deal as well for making implicit memories stronger. I haven't explored that but it's probably key. Addition:

It seems like it would be political suicide, but so did trying to coup the government. The fact of the matter is, the citizens are too caught up in their own lives to care about a nazi salute or a government coup or a convicted felon with a several decades long history of con artistry . Half of them probably don’t even know what a Nazi salute or a Nazi is. The majority don’t care but the vocal minority opposition does, which is who the salute was intended for.

Right, the Nazi salute isn’t surprising to me, it’s just clear evidence for those still pondering. Lol

Doesn’t matter if he’s a nazi or not. He did a nazi salute during the presidential inauguration and it was applauded and then hand waved away by nearly all in attendance. Doesn’t matter how immature or dumb he is. This is a testament to the lack of intelligence, integrity, and common decency of this new administration and its lieutenants. He did the Nazi salute because he’s emulating Trump’s playbook. Do some wild shit to get press and polarize people on your behalf. Make people defend you so they entrench themselves in your corner, which effectively creates a cult-like following. He will do more and more things like this, and it’s not because he’s simply against the woke mind virus. He’s using that as cover for gaining a following. People think he’s more ideological than what he is. He wants material gain and power. He doesn’t give two fucks about a woke mind virus. He would embrace the woke mind virus if it gave him power and material gain. It’s just his tool. This is obvious now. He doesn’t care about his trans son or humanity. He’s not heart broken about anything. Ta da! Now you know Musk.

I thought about it. I'd be down for it but I don't like dealing with clients. It can be a pain in the ass. If someone was willing to do all the sales and client/account management, I could get on board with it. Also, maybe a better idea is to come up with complete business solutions for tiny businesses. For example, there are tons of successful one-man and even 3-5 man HVAC businesses that are stuck and don't know how to scale. You could offer them a complete brand, digital, and business strategy package for like 40-50k a pop and have all the systems basically plug and play. It would take a bit of set up in the beginning but once all the systems were dialed in, it would mostly just be a matter of sales. I've heard of people doing things like this and it seems like a sweet ass gig.

If you know HTML, CSS, and basic JS, you should be able to use Claude and ChatGPT to build it. If you don't know those, you're probably a good ways out before you could produce a solid web app. And if you decide to get into modern frontend tools like React, Tailwind, etc, you'll have an even steeper climb. If you're a newbie, the problem is, you don't know what you don't know. It's very easy to build something that works but it's hard to build something that works well at scale, because to do that, you have to build things right, and there are dozens of ways you could build this app. Your entire tech stack would need to be well mapped out and understood. Such an app would need complicated state management solutions, so most would tell you to use React. I don't like React because it seems overly complicated, so I decided to build my own solutions for rendering elements in the DOM and managing their state. I plan on eventually building commercial apps myself, but first, I'm setting up a "systems builder" desktop app (screenshot below) that lays out my entire architecture with documentation and easy copy/paste for all my code. I'm creating all the building blocks I would ever need, that way, when I'm building my apps, I'll have everything ready to go. I'm using custom web components, which is where you build out your own custom elements and program them to behave however you want. Say you want to have a file drop element where people can upload a profile picture. Well, you could build it out normally with HTML, CSS, and JS, but it's not very efficient for reusing it, so you can put it all in a custom component and make it work however you want. I've designed my file drop component to work like this: Now, anytime I want a file drop element, that's all I have to do. The idea is to build things so they can be used with as little friction as possible. That little bit of HTML sets up the functionality you see above, which is actually somewhat complex behind the scenes. And, to make sure I never forget about it, I'm setting up documentation for each component, which provides a demo and code for easy copy and paste. Once this system is set up, I'll be able to scaffold out apps very efficiently. It's taken me a while to figure out the best way to set this all up because I kept realizing I was handling things suboptimally, so I had to go back to drawing board and start from scratch many times. I didn't know what I didn't know. I'm finally at a point where I feel good about the architecture and everything but it took a lot of time. So if you have the money and don't care much about this tech or acquiring the skills, it's probably better to just pay someone, otherwise, there is a good bit to learn and the only way is to get your hands dirty and keep them dirty for a couple years.

You misunderstand what current AI is. LLM stands for large language model. The whole point of it is to predict it’s next output based on a given input. They trained it on language and had it make millions of guesses and the programmers told it if the guesses were accurate. Over time, it’s predictions got better and better until eventually, it could learn unsupervised. Im not the best to communicate it but AI is not programmed the way a calculator is. That would be simple. If you think an AI is the same as a calculator but just scaled up for every possible question, then you will be shocked when you learn what it actually is. They don’t even know how it does what it does. They can’t trace all its processing. They call that “the black box”. It literally is like a giant mind. It wasn’t trained to produce a specific output, it figures it out on the fly, guided by parameters of course. For example, they have it behave friendly. They did program that in. Look up neural nets, machine learning, and natural language processing, or better yet, just ask AI to explain the mechanics of itself. # Mechanics of AI Large Language Models (LLMs) ## Architecture - Based on Transformer architecture introduced in 2017 (Vaswani et al.). - Key components: - Encoder-Decoder structure (e.g., BERT). - Most LLMs like GPT use only the decoder. - Utilizes self-attention mechanisms to process context. ## Tokenization - Text is broken down into tokens (words, subwords, or characters). - Each token is mapped to a numerical vector using embeddings. ## Training - Trained on massive datasets containing text from books, websites, and other sources. - Objective: - Predict the next token in a sequence (causal modeling). - Fill in blanks (masked modeling, e.g., BERT). - Loss function: Measures the difference between predicted and actual tokens during training. ## Context Window - LLMs have a fixed "context window" that limits how much text they can process at once. - Larger context windows allow processing of longer texts but increase computational cost. ## Attention Mechanism - Self-attention allows the model to focus on relevant parts of the input sequence. - Calculates relationships between tokens to understand context and semantics. ## Layers and Parameters - Consist of stacked transformer layers, each containing: - Attention heads. - Feed-forward networks. - The number of parameters (weights) determines model size and capability (e.g., GPT-3 has 175 billion parameters). ## Pretraining - Models are trained on general tasks using unsupervised learning. - Enables them to learn patterns, grammar, facts, and reasoning. ## Fine-tuning - Optionally adapted to specific tasks (e.g., sentiment analysis, code generation). - Done using supervised learning on domain-specific data. ## Inference - At runtime: - Given an input prompt, the model predicts and generates the next token iteratively. - Continues until it forms a complete response. ## Scaling - Larger models tend to perform better but require significant computational resources. - Scaling laws predict improvements with: - More data. - Larger models. - Longer training times. ## Limitations - Prone to generating plausible but incorrect or nonsensical answers (hallucinations). - Lack common sense and grounding in real-world truth unless externally validated. - Memory is limited to the input/output within the context window; lacks long-term memory. ## Applications - Can perform tasks like: - Text generation. - Summarization. - Translation. - Coding. - Question answering. - Used in industries such as education, customer service, and research.