My point thoughts as I read: that's it, it's right on his shirt, the future will not be supervised, it is all about the next big thing in the AI field - Unsupervised Learning......physics does it's thing on its own and this is it
yes OpenAI staff are smart, made/improved (don't we all!?) big things, and are safety nuts, which gains max love from the simpleton human race, half of which is called for. I see why some my friends say they have a distate of openAI, but I don't see it that way, they did lots of good. GPT-2 sealed it for me P.S. no AI isn't brittle, only smaller are more brittle, intelligence=Good not evil, you can't have an evil/brittle intelligence since a large aligned brain/team are only a shared context powerhorse WHEN aligned. Those GAN adversarial attacks are just stupid AI, human brains aren't fooled by them. Right there you have transparency with better AI, see. No issues they do want money probably, so their image and safety stuff is probably over hyped, but they are doing lots of good and spreading the word, I'm very pleased what is this, the article says inside their place they argue/are unaligned, with no proof or this claim... omg, i never expected their place to be a total dump, wow INFINITE JEST, WHAT!!!?? WHAT THE HELL IS THAT!!? YOU CAN SEE IN THE IMAGE A SUPER PRETTY BABE IN A BEAUTIFUL JACKET IN A SMALL LUXURIOUS ROOM, AND THERE IS A SIGN THAT SAYS 'NO FOOD NO DRINKS, OR ELSE WE KILL THE W...' O.O indeed they appear open to save the world and appear to have 0 pay (but, they do! right!?)....but there's no proof of evilness here still! The article 'says' they (which seems plausible) went nonprofit (even tho have 1 billion$+) to bring n the cash all while looking Good, and indeed that could all be true that they will share all their work and work hard at AGI....and yes they got another billion+ and a new for profit but maybe they really are trying and going to share (or sell?) their work etc. Yes mass trust would be the goal, to get most friend network connections flowing open channels to pump them big the open AI portraits on their wall......HAHAHAHA so funny ya they profited perhaps off their staged GPT-2 releases (which was amazing technology), they said it was dangerous maybe yet release it to us, while looking like they save the day for being safety experts, and then top it off by saying it was a test/show to other researches how to introduce security. Then say they'll slowly be less open/publish so to Follow the charter, that's a trade off of distributed work - less work but more safety but remember I said safety comes from mass sharing. But they will eventually share as they already are as well doing anyway. There's really nothing wrong here, they are doing a lot of good even if are money hungry. I see they have a few domains working on different areas - robotics team, language team, indeed, and as time moves on they will unite aligned to work on a clear specific AGI path, and you bet it will be a 'language' path, as it all is that, data is language, a story of physics diversity in color, i do not care if they have 20% females, thats the way life is in their region who cares I see they found out big data and upping others creations works, and indeed this is how all inventions are made! I suggest indeed they hide it, profit, then share, that's natural. And omg, yes we want them to have lots of resources, it costs us nothing and gives us a huge change of becoming immortal by AGIs. yes... "hype in its rush to attract funding and talent, guards its research in the hopes of keeping the upper hand, and chases a computationally heavy strategy—not because it’s seen as the only way to AGI, but because it seems like the fastest." ....who would pass up the right to do so....and they are sharing it all eventually more and more as well anyway "OpenAI needs to make money in order to do research—not the other way around." ....Well I don't need any lol really that was hilarious openAI read this blog dump post and the writer talks about their next big project we're waitin for ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/Tab97f0f82cc3442a-M97e5327771888723f79c2a82 Delivery options: https://agi.topicbox.com/groups/agi/subscription
