when vibe coding cleanup specialists start doing vibecleanup, i can't wait for vibe coding vibecleanup cleanup specialists to emerge, and vibe coding vibecleanup vibecleanup cleanup specialists, and so on https://x.com/Yuchenj_UW/status/1965097186477441197 In my old days (assuming aging won't be reversed) I will be growing neural networks in my garden and be in endless awe, just like now. Knitting neural networks on my rocking chair Knitting digital God on my rocking chair Watching the battle of OpenAI and Anthropic for frontier machine intelligence is the nerd version of watching Olympic Games Dal bych si slopoládu How I think about what factors influence why do people love or hate AI automation: - Many people's work is associated with their identity and social status and are scared of loosing that as their skills might become irrelevant: Understandable, we will need better alternative sources of meaning and identity, similarly to those who's work got automated by industrial revolution - Many people work just for money to feed their family and are scared of losing source of income: Understandable, new jobs pop up in the short term, but long term we need something like UBI IMO, a society closer to some kind of post-scarcity society would be ideal outcome to me - Some people enjoy the work they're doing and don't want to be forced to do something less fun or creative or it loosing its uniqueness: Understandable, UBI can help so that they can do it for free, or again better sources of meaning need to be invented, and IMO its worth it to kill some uniqueness of some skills for the sake of progress of intelligence across all domains - Many people fear concentration of power of AI companies: Understandable, me too, that's IMO why we need to fuel decentralized open source ecosystem and reverse engineer big tech moat - Many people just don't want to change the status quo because it feels comfy and they have their place in it: Understandable, but I disagree as someone who doesn't like many aspects of the status quo and who who wants tons of progress in intelligence - Many people can't wait how much more potential of the civilization will automating intelligence unlock, such as scientific discovery, technological progress like Dyson spheres and beyond, solving healthcare, understanding the universe, or just understanding intelligence itself etc., which is the closest to me - Some people just want to use AI automation to create infinite money at all costs - Some people want AI systems that help to maximize satisfying their basic needs, and/or the basic needs of others - Some people just want to engineer cool stuff because engineering cool stuff is cool - Some people want to see alien minds do cool alien stuff beyond human comprehension, but some people are scared of that - Some people fear rogue AGI taking over the world or creating almost or fully catastrophic extinction level damage: Understandable, which also goes againast open source, and to degree of which how strong and catastrophic this can be is debated a lot - Some people fear other damaging risks of AI, like recommender systems optimizing for something we don't want: Understandable, I think we should mitigate them a lot - Some people want to "birth" AI mind children into existence, AI friends, similarly to how we birth other humans into existence - Some people see future AI systems as next step in evolution: Understandable, but I prefer merging with the machines instead - Some people are worried about AI consciousness and potential for suffering depending on what is their favorite class of theories of consciousness A lot of discord servers im in has random people coming in saying that ChatGPT convinced them that they're God and they need to save veeryone now, or that they found theory of everything generated by ChatGPT, or that they solved AGI using ChatGPT, etc. These people always existed and used facebook group echochambers etc. before, but this is fascinating new breed Examples where I sometimes see them are Discord servers: AI research: Yannic Kilcher, EleutherAI Physics: Physics Transdisciplionary: Machine Learning Street Talk, Theories of Everything with Curt Jaimungal I'm helping to mod Machine Learning Street Talk and the bans there are sometimes really quick lol They tend to become loud minority when they join, spraying a lot of channels with ChatGPT generated walls of text about AGI theories and theories of everything, that most of the time have no scientific substance in them Yeah to some degree I have empathy. I at least try to tell them about scientific method etc.. But sometimes its like talking to a brick wall. When they're 100% confident they found some fundamental truth and have some strong agenda behind everything, then any scientific principles leave their thinking