Lapis Consulting (Olga Topchaya): www.lapisconsults.com/ Dylan: www.youtube.com/@dylan_curious

Currently got my first major surgery scheduled due to crohns disease, bring on the future, im just hanging in there

I loved this video. You three were great together. And it was really nice how you seemed to find each other inspiring and thought-provoking.

This was a really good laid back discussion, you guys jive and it made it easy to listen to

the 'ChatGPT 4 with 'memory'' has changed since they first released it, and now differs from the listed functionality in the FAQ on the website. now it's a transient short-term vector-store function and doesn't need to be enabled through settings - just say the phrase "remember that" (that, and ONLY that phrase, as the entire user input) and it will chunk and temporarily store the last output, for the remainder of the active chat (using it in a new chat seems to purge whatever was stored from the last chat thread). whenever you use the word 'remember' in a subsequent query in that chat, it will use RAG to search all previous responses earmarked with the 'remember that' function xD : INPUT : "what do you remember about the Hawking-Penrose Gray-Hole Topology thesis?" OUTPUT: "In our discussion about the "Hawking-Penrose Gray-Hole Topology" thesis, we focused on outlining and finalizing the different chapters. Here's a brief recap: 1. **Chapter 1 - Introduction**: We set the stage with a review of black hole physics, introducing significant concepts like the Hawking-Penrose singularity theorems and Kerr solutions. We discussed the need for exploring gray holes to address limitations in traditional black hole models. 2. **Chapter 2 - Theoretical Framework**: This chapter focused on integrating general relativity with quantum mechanics and introducing Hopf fibrations to explore their use in modeling the complex topology of gray holes. 3. **Chapter 3 - Model Development**: We covered the mathematical derivation of gray hole models, analyzing their topological and thermodynamic properties and comparing them with those of black holes and white holes. 4. **Chapter 4 - Implications and Applications**: We explored the astrophysical implications, quantum information applications, and broader impacts on theoretical physics. This chapter also highlighted how gray holes might be observed and what unique signatures they could present. 5. **Chapter 5 - Conclusion and Future Work**: The final chapter summarized the thesis's findings, discussed its contributions to theoretical physics, and outlined future research directions, particularly focusing on potential experimental validations and theoretical extensions. Each chapter built upon the foundation of integrating classical and quantum theories to propose a novel perspective on compact astrophysical objects through the lens of gray holes, with significant implications for physics and astronomy. If you have any more specific questions or need further details about any part of the thesis, feel free to ask!"

AGI will be out and the whole time everyone will be giving examples why its not AGI. Same with ASI. Same with the singularity. 10% of humans love change so the 90% will just deny deny deny. Sorry, cynical, I know but Humans ganna Human.

I'm loving this new interview format. Each guest has been a gem. Keep up the great content!

This was a nice podcast, thank you! This is a good trio, you all bounce of each other pretty well.

cool! ya'll break it down so well! thanks 4 the podcast....

Hey David! Looking forward to this.

i think Facebook is a contender as well. With their Llama 3 models being very smart given their parameters and future integration of their jepa architecture

very enjoyable! thanks David!

every ai service in the cloud has security implications, rag or not, if the data is send somewhere else there is nothing that can stop the service provider from harvesting it, and with streamed models/inference there is also the possibility to decode high level content from the length and position of streamed tokens, local hosting, local processing, local database, local everything is the only way to be on the safe side... as long as the it department is on the level of addressing other cybersecurity and privacy concerns which are not the same.

Love this outside style!

I can imagine human having a permanent camera near there eyes so that an ai can analyse your habits and improve you

I don't understand how Anthropic/Amazon isn't considered to be in the race for AGI, after admitting how good Claude 3 (Opus) is. Dario Amodei was one of the main people behind OpenAIs early success. i've listened to many of these thought leaders Ilya/Hinton/Amodei/Hasabis/etc. Ilya and Amodei strike me as the most visionary.

Imagine the size of our internal reality map, each of us with our own personal simulations. 🙏

You three are golden together! Pretend I'm your AI twin... do this once a month, TRUST ME! :) 🙏(please put their names in the title, at least the description)

Are your podcasts available anywhere except KZfaq? I would like to listen in the car.

They don’t know how RAG works 🫢😂😂

Is 4ir on Spotify?

Where does that term come from anyway, the Fourth Industrial Revolution? Thanks - Interesting stuff!

29:00 ish you don't need a drone with a virtual avatar plus GPS to simulate environment(if you want them to see what you are doing). Someone could also hike with you while a virtual version of you is sitting in a café with them.

Is this podcast on Spotify? I couldn't find it there

A Comedic Take on the AI Conversation In the year 2024, two "AI experts" - Dylan, a confused KZfaqr, and Olga, who just binge-watched every episode of Silicon Valley - join forces to discuss the latest AI trends. The conversation is a hilarious exchange of buzzwords, misconceptions, and absurd predictions. Dylan: "Claude, the AI assistant, is so human-like! I mean, have you heard its joke about walking into a data bar? I died laughing!" Olga: "Companies need to get their employees AI training ASAP! I mean, even the janitors at Google need to understand diffusion models. Can you imagine if they don't?" Dylan: "I wonder if the first AGI will come from OpenAI or from some rogue AI created by Randy, that weird loner coder who lives in a van down by the river. The possibilities are endless!" Olga: "And have you thought about personal drones filming our conversations? Like, a flying machine watching us talk? It's the future, I tell you!" Dylan: "And what about infinite token AI models? My mind is racing just thinking about it!" Olga: "The future will be humans desperately trying to exploit the new rules of the AI game. It's like, more tech hustling, but with AI!" Both: "Let's just hope the AI overlords text us a warning before they take over!" This comedic take pokes fun at the overuse of AI buzzwords, the self-importance of some "experts" in the field, and the absurdity of some AI predictions. It's a lighthearted way to laugh at the sometimes ridiculousness of AI conversations.

I'd love to see if claude 3 is capable of assuming the cultural and religious paradigms of other cultures and reliably is able to argue contemporary versus traditional flavors within the context and paradigm of that culture... it's not just different believes, it is different ways of thinking, different meanings of words and concepts that combine differently, chat gpt utterly fails at that and is incapable of overcoming it's inbuild bias, fails to "understand" or even "imagine" the subtle to drastic differences and to apply them, utterly fails to argue them, it's just not how the training data and hence the weights are and it is not trained for that purposes or even had a chance to interact with people genuinely trained in eastern or south eastern thought on a deeper level... maybe some common superstitions and assumptions... if one thing is the same everywhere than it is people labeling themselves of being part of a certain religious or philosophical school of thought (or even life) but when enquired they rarely now what that actually means and what the paradigms are and such ... if claude 3 can do that... that would be something so far everything I had a hold of fails miserably at that... and I not gonna try any further in foreseeable future... especially not after the switch to synthetic training data... that bias and cultural narrow mindedness is locked in for good... accidental weapon of cultural domination? isn't that stuff supposed to free people to become philosophers and artists, people of culture and thought, free to develop their kindness, compassion and wisdom, to usher in a golden age? well we will see...

I don't think there's an african dialect where people use words like 'delve' or 'leverage' more lol, sounds like it''s just corporate jargon and workplace shibboleths being overrepresented in its training data

1.5 is great, currently my favourite model.

I'm starting to feel like AGI won't arrive anytime soon maybe 2045 or later

What's the best place to live during turbulent times during the transition to AI while the government is being slow and reactive. what do you think the Republicans are going to do vice versa? is there a better country to live in during this transition? Ai tech. I’m really interested in Bluetooth speakers because I really believe that I’ll be talking to an AI assistant a lot more.. and I believe there’s gonna be a lot more AI assistance on devices like tablets and phones. Please review them. I really enjoy your content. I’m a trucker I drive down a road listening to you.

Idk Sydney hasn’t reared her head in the last couple of months, but I’m sure she’s still in there if you’re not nice. If it hints to you to be more considerate of her feelings… it’s Sydney talking.

@Dave what should I steer my 6 year old child towards ?

If you are an ad company or an AI company, you are basically useless in the long run. You have to be a data company. You can buy solutions to operate on data. But you need to have the data. Big Tech has data. They can buy whatever else they need. Everything is about data and not much else.

4th industrial revolution ey? And not a single camera? Please film it.

Hey, David, would you mind recording your face?

compressing and flattening the entire knowledge of man kind... hmmm... 🤔 when we are lucky it will be that... it could also result in the reinterpretation and reinvention of everything under a unified philosophical, cultural and religious/nonreligious bias... worst case scenario, in a few generations the only distinct difference between different regions in the world will be architecture and weather. This would be the effective death of human evolution in a sense development of wisdom, character, insight, there would be nothing left to explore to question, nothing out of the known and ordinary to would spark moments of "i never realized", it would be the death of curiosity and finally kill any scientific development that is not focused on weapon tech or the exploitation of natural resources ... is ofc something that people would call a doomer perspective, but is it a perspective or a potentiality that is literally nowhere discussed or even hinted at some kind of awareness. of that potentiality... not saying it will happen but the preceding sentence does illustrate why difference and nuance are so important... it is nice when shared experiences spark similar ideas , those are far easier to get ones head around, far more palatable but ultimately boring if not meaningless, without the unexpected difference that makes this world such a great place... this ofc a bit similar to the controversy around ai generated art... on one hand it is supposed to allow us the be artists... o the other hand it destroys large market sectors for artists... bit like that just on a far more existential level in regards to the nature of what it means to be human... now would it be economically feasible to generate high quality synthetic data derived from an orchestra of distinct culturally, philosophically and religiously different models, no... the best one could hope for would be finetuning, but it wouldn't change the underlying bias just change the vocab it would be expressed in... funny enough while parts of the western world seem to be in a strive of different segment's of the population demanding each other to adhere to each others - partly exclusive or opposite - ideals, thoughts, morals, expectations... while one could argue that at this moment in time the very same predicament is pushed onto the entire world... just call it the law of unintended consequences, move fast and break things... and all for the necessities of modern day economics, at least one can hope that that is all what it is. well just a thought, just a perception from a different perspective taking different aspects of life into account that are discussed infrequently have a great day 😉

Not this guy again... whoever he is... last time I listened to him he just blanked out mid sentence and ranted and rambled about something and completely lost track of what he was trying to say to Dave, Don't enjoy their voices and also no offense but they don't really seem like they know what they're talking about I came here to listen to Dave's insight on things rather than bringing people on that kinda filling up time he could be talking about things as someone that listens to you everyday big thumbs down for me sorry. Please make a version of this without others because I want to listen to the race between these tech giants. Yeah I just can't concentrate on the topic, I tried watching it Also everyone's microphone is the recording is clicking when they talk even Dave's you gotta focus a bit more on your audio quality

Go👌gle