Wisdom in the Age of AI: A Philosophical Quest with Vervaeke, Pageau, and Schindler

DC Schindler explained a lot of how I feel about AI way better than I ever could.

The AI explosion last year is the reason I found this wonderful channel. For that I'm thankful.

"Idol Problem." Pageau is spot on. Always remember what he said.

Wow, once again I am astonished by these conversations. They are so timely and needed. If the apple vision pro becomes widely used, I'd hack it to make people watch this content.

Excited to see this! John, have you seen David Shapiro's essay on moloch and the meaning crisis? He goes over some of your work as well as Boeree's and Schmachtenberger's, and relates it to his own in the sphere of machine learning and AGI thresholds. He's right on top of the latest technical developments in this realm, and I think he'd be a fantastic addition to the conversation 🙏💕

Ask yourself this question. Would Mr. Vervake still litteraly be the top mind in the world explaining this stuff. If he didn't understand Plato? 😳🤣... So wild. 😆 I love this man. In the philias sense of course.

This was a fascinating conversation!

Just saying: You have incredible insights!!!

What an important discussion! JVs strategy of demonstrating the best of humanity to AI seems an ideal goal. But we need to be cognisant the AI is less like an emerging intelligent being and more like an emerging nation, akin to a sovereign jurisdiction (subject to the scale of the AIs parent entity and its relationship with 'Global Standards and Rules').

I think Dr.Vervaeke and the fellows talk is helpful to understand the property of AI development and IT programmers also find learning from this talk. ChatGPT's developed company OpenAI also has committee members who have same line of opinion as Dr.Vervaeke told, it means IT programmers can have conditional reductionist view, and many programmes need to listen up the opinions. The majority of people have same common sense to development of technologies. I am sorry if my reply is becoming long length and end up this reply ^^ I learn much from Dr.Vervaeke and his book about AI, I am going to get to read all ^^

I always wonder who "we" is. Most people don't see the black boxes or the genies flitting about. Some investors might. Oh, it's Mammon, alright. AI will reflect Sapiens, and that's that. If God possesses AI, well, that would be a miracle. "I'm going to win...?" WOW! How many Indians and Chinese nationals are in the audience? This was a very intoxicating conversation with many points that align with my intuitions. Something will emerge that we can't predict. It's a fascinating subject. #influenced Who is David, I really resonated with his insights.

Thanks for this John! This is great, it's a bit different, but on the same vain as LLMs, I tried Tesla's FSD (Full Self Driving) this week and I was blown away. I saw the future arrive and I could see it's limitations and issues on multiple levels (and has been keeping me up in the middle of the night!). This is also a game changer. Love this conversation, it is essential. (We need to get Elon here 😁)

Grog uses LPU's “According to Groq, in similar tests, ChatGPT loads at 40-50 tokens per second, and Bard at 70 tokens per second on typical GPU-based computing systems. Context for 100 tokens per second per user – A user could generate a 4,000-word essay in just over a minute.” What is LPU, how does it work? ChatGPT The term "LPU" can refer to different concepts depending on the context in which it's used. However, in the realm of technology and computing, "LPU" isn't a widely recognized acronym with a single, definitive meaning. It could be a typo or confusion with other more common terms such as "CPU" (Central Processing Unit), "GPU" (Graphics Processing Unit), or even "TPU" (Tensor Processing Unit), each of which refers to specific types of processors designed for different computing tasks. CPU (Central Processing Unit): The primary component of a computer that performs most of the processing inside a computer. It executes instructions from programs, performing basic arithmetic, logical, control, and input/output (I/O) operations specified by the instructions. GPU (Graphics Processing Unit): User language Processing Unit what is it how does it work? ChatGPT The term "Language Processing Unit" (LPU) is not widely recognized as a standard component or technology in the field of computing or artificial intelligence as of my last update in April 2023. However, the concept can be interpreted or imagined based on similar technologies and principles used in natural language processing (NLP) and related fields. If we consider an LPU in a hypothetical sense, it would be a specialized processor or computational unit designed specifically for handling natural language processing tasks. Hypothetical Design and Functionality An LPU, in a speculative sense, would be optimized for algorithms and operations commonly used in NLP, such as: Tokenization: Breaking down text into words, phrases, or other meaningful elements. Part-of-Speech Tagging: Identifying parts of speech in text, such as nouns, verbs, adjectives, etc. Named Entity Recognition (NER): Identifying and classifying key elements in text into predefined categories, such as the names of persons, organizations, locations, expressions of times, quantities, monetary values, percentages, etc. Sentiment Analysis: Determining the sentiment expressed in a piece of text, be it positive, negative, or neutral. Language Translation: Translating text from one language to another. Speech Recognition and Generation: Converting spoken language into text (speech to text) and vice versa (text to speech). How It Could Work An LPU would likely incorporate specialized hardware accelerators optimized for matrix and vector operations, which are fundamental to the machine learning models used in NLP tasks. This could include: Custom ASICs (Application-Specific Integrated Circuits): Designed to accelerate specific types of neural network computations with high efficiency. FPGAs (Field-Programmable Gate Arrays): Which can be reprogrammed to suit different tasks and algorithms, providing flexibility alongside performance. Optimized Memory Systems: To handle the large datasets and models common in NLP, an LPU would need highly efficient memory management systems to quickly access and process data. Current Analogues In the absence of a widely-recognized LPU, tasks related to natural language processing are currently handled by general-purpose CPUs, GPUs, and more specialized processors like Google's TPUs. GPUs, in particular, are well-suited to the parallel processing requirements of machine learning and NLP tasks, offering a balance of flexibility and power. Conclusion While the term "Language Processing Unit" suggests a fascinating direction for the future of computing hardware, especially tailored to the demands of natural language processing, as of now, it remains a speculative or theoretical concept. Advances in hardware specifically designed for AI and machine learning tasks continue to evolve, and it's possible that we'll see more specialized processors like an LPU in the future.

As a longtime professional working in machine learning and artificial intelligence, I can tell you this level of dialogue is sadly almost entirely non-existent. This is much appreciated, everyone.

🙏

Ironically...it requires significant intelligence...to achieve an understanding of...intelligence. Which begs the question..."what is understanding?"

Who's telos are we satisfying ? --Implicitly power issues emerge --ie for the few or the many ?

I struggle with intelligence and understanding. If an LLM doesn't understand what it's doing how can we even say it is intelligent let alone rational

❤

~15:00 theology point