@CMcdazzle @pleovision @alcuintweets @chatokox3 yeah I just need to hook that up tomorrow, been busy with work

@guerre_doctrine yes... yes we should... this stirred a deep memory I forgot I had... how do you do it...

@guerre_doctrine gem so bright I went blind

@XorDev I thought Inigo had peaced out from developing it, glad to see it's back. I've been a happy fragcoord user though

@punished_teno thanks for buying me one too 🙏

I've always believed that intelligence should come from simple, general building blocks that learn their own abstractions directly from raw data; in other words, the bitter lesson. I think we're finally seeing a real step towards AI that can learn from all data.

Language s a microcosm of this ideal, and to me it's a central problem in AI: if we can't build LLMs that learn end-to-end from bytes, how will we handle the rest of the world's data that we understand even less?

Last year we introduced hierarchical networks (H-Nets) as a first step. Our new version finds up to 3-5x efficiency gains in controlled scaling law experiments, and at larger scale matches an open-source model trained on 11T tokens with ~10x less data and compute. The posts below walk through what we found and why we think it works.

単位を変える魔法

@wavefnx I jest of course

@wavefnx it's too horny to be pure sweet Maschinenmensch Maria, the Sorayama neuron is too strong

@wavefnx hajime sorayama

@skooookum Charlie is ready for war

@sameQCU Thank you same I’ll shop around

@sameQCU What would QCU Labs recommend

@cosmojg @pleovision @kojinglick I really need to give it full thread context

@thinkingshivers cool! I played way too much enclose horse

@norvid_studies what does that even mean Norvid

@hyrevayn I should note that this is good looking simply because I'm optimizing for speed

@kojinglick the question comes by way of people telling me prices are doubling and will soon be tripling

why would I do such a neolithic thing as editing pixels by hand? Well sometimes I need the tiniest of edits that are faster like this.