3 ms·

ChatGPT hasn't overcome any of the fundamental issues, it's just a huge improvement on the things that the original GPTs were good at. Being able to stay cohere

ChatGPT hasn't overcome any of the fundamental issues, it's just a huge improvement on the things that the original GPTs were good at. Being able to stay coherent for a trained-in length that gets longer with larger models is different from the length-unlimited coherence that human beings can manage, spanning lifetimes of thought and multiple lifetimes of discourse.

dragonwriter 4y ago
> ChatGPT hasn't overcome any of the fundamental issues, it's just a huge improvement on the things that the original GPTs were good at. Being able to stay coherent for a trained-in length that gets longer with larger models is different from being internally coherent for 90 years by nature like people are. People are very often not internally coherent over periods much shorter than 90 years.
- whatshisface 4y ago
  The hazily defined ideal of general intelligence - which everyone imagines that they are, but most of us do not live up to all the time, and nobody lives up to every waking moment (or at least before we've had our morning cup of coffee) isn't within the reach of present day transformer architectures because the length of text they can stay coherent over is limited by how far back they can look through their own output. Human beings can form new memories from their outputs that stay with them for the rest of their lives.
IIAOPSW 4y ago
>length-unlimited coherence that human beings can manage I've yet to meet a human that doesn't eventually run out of steam. You just aren't well situated to notice decoherence on time scales longer than your own coherence.
nl 4y ago
In 2016 Transformers didn't exist and the state of the art for neural network based NLP was using LSTMs that had a limit of maybe 100 words at most. With new implementations like xformers[1] and flash attention[2] it is unclear where the length limit is on modern transformer models. Flash Attention can currently scale up to 64,000 tokens on an A100. [1] https://github.com/facebookresearch/xformers/blob/main/HOWTO.md https://github.com/facebookresearch/xformers/blob/main/HOWTO... [2] https://github.com/HazyResearch/flash-attention https://github.com/HazyResearch/flash-attention

dragonwriter 4y ago

> ChatGPT hasn't overcome any of the fundamental issues, it's just a huge improvement on the things that the original GPTs were good at. Being able to stay coherent for a trained-in length that gets longer with larger models is different from being internally coherent for 90 years by nature like people are. People are very often not internally coherent over periods much shorter than 90 years.

whatshisface 4y ago
The hazily defined ideal of general intelligence - which everyone imagines that they are, but most of us do not live up to all the time, and nobody lives up to every waking moment (or at least before we've had our morning cup of coffee) isn't within the reach of present day transformer architectures because the length of text they can stay coherent over is limited by how far back they can look through their own output. Human beings can form new memories from their outputs that stay with them for the rest of their lives.

IIAOPSW 4y ago

>length-unlimited coherence that human beings can manage I've yet to meet a human that doesn't eventually run out of steam. You just aren't well situated to notice decoherence on time scales longer than your own coherence.

nl 4y ago

In 2016 Transformers didn't exist and the state of the art for neural network based NLP was using LSTMs that had a limit of maybe 100 words at most. With new implementations like xformers[1] and flash attention[2] it is unclear where the length limit is on modern transformer models. Flash Attention can currently scale up to 64,000 tokens on an A100. [1] https://github.com/facebookresearch/xformers/blob/main/HOWTO.md https://github.com/facebookresearch/xformers/blob/main/HOWTO... [2] https://github.com/HazyResearch/flash-attention https://github.com/HazyResearch/flash-attention