3 ms·
Google pumping out PR to get some damage control on chatgpt. No evidence in the public of any of these claims. Show us the models google
by greatpostman 4y ago
Google pumping out PR to get some damage control on chatgpt. No evidence in the public of any of these claims. Show us the models google
- we_never_see_it 4y ago[dead]
- nightski 4y agoLecun @ Meta has been doing the same thing on Twitter lol. Re-tweeting all these papers with no actual public information.
- waboremo 4y agoHalf of the reason I still have a twitter is for this very reason. Highly intelligent people exhibiting some of the most petty behavior. It's a great reminder, of many things, including how tribal we still are in the worst ways.
- amf12 4y agoThese are based on published research by Deepmind, which is an AI research lab. They don't build products to "demo" anything to the public. They research and publish peer-reviewed papers. The only thing they can "demo" is the simulation of the model, which they have published as videos on YouTube. Here's the paper if you're curious btw: https://arxiv.org/pdf/2301.07608.pdf https://arxiv.org/pdf/2301.07608.pdf
- greatpostman 4y agoAll I hear about is deep mind, when will they release a model I can interact with? That’s the standard chatgpt set
- jjfoooo45 4y agoPlenty of academic research in computer science comes with concrete demos of capabilities