LLMS

→ View original post on X — @marktabnet

GitHub Launches GPT-4-Powered Copilot Coding Assistant

By

@marktabnet

–

28 March 2023 22h10

GitHub launches its new GPT-4-powered Copilot coding assistant https://
bit.ly/40FPnW2

→ Voir le post original sur X — @stephanemallard

AI May Destroy More Jobs Than It Creates, Goldman Sachs Warns

By

@stephanemallard

–

28 March 2023 21h21

L'IA créera énormément de valeur et beaucoup de nouveaux emplois (encore inconnus aujourd'hui) mais peut-être moins que ceux qu'elle détruira (destruction nette) parce qu'elle s'attaque à la dernière spécifié de l'Homme au travail : l'intelligence https://
lesechos.fr/tech-medias/in
telligence-artificielle/chatgpt-et-lia-menacent-300-millions-demplois-dans-le-monde-selon-goldman-sachs-1919968
…

Ideas for Improving Language Models Without New Developments

By

–

28 March 2023 19h27

reply if you have other ideas about how language models could currently be improved without any new developments and I'll add them to the thread

Language Models Gaining Access to Tools and APIs

By

–

28 March 2023 19h27

giving language models access to tools and api’s chatgpt has done this with plug-ins, no longer can you say chatgpt is bad at math or doesn't have access to current info soon every api on the internet will be forced to have language model interaction documentation

Language Models with Long-Term Memory via Embeddings and Search

By

–

28 March 2023 19h27

language models with long-term memories using embeddings and semantic search language models will be able to access some short convo you had a month prior

Chained prompt sequences as the future of AI APIs

By

–

28 March 2023 19h27

using pre-created chained-prompt sequences instead of just one single prompt as api costs continue to decrease this will become more popular, the intermediate prompts will be hidden and only the output will be shown could imagine chatgpt integrating a "prompt marketplace"

GPT-4 Vision Capabilities: Multimodal Reasoning and Image Generation

By

–

28 March 2023 19h27

I haven't even touched on the vision capabilities possible in gpt-4 some obvious ones:
-multi-modal reasoning and action (e.g. MM-REACT paper)
-better image gen (imagine gpt-4 iterates on your Midjourney prompts) -few-shot prompting and learning through images/videos

Unlocking Language Models’ Potential Without New Development

By

–

28 March 2023 19h27

gpt-5 is not needed to 100x the potential these models have we could stop all language model development today and we still haven’t scratched the surface of their capabilities here are a few non-obvious ways language models can be improved without creating any new models: