Pedro Domingos quips: 'new idea' called RNNs will power next-gen LLMs
pmddomingos · x · 2026-09-05
ML scholar Pedro Domingos posted a tongue-in-cheek take: "I have this amazing new idea called recurrent neural networks. They blow transformers away if you do them right. The next generation of LLMs will be based on them." The sarcasm nods to the ongoing revival of recurrent-style architectures (linear attention, state-space models) in the LLM research conversation.
More from Fun
- METR staff finish eval sprints, open Twitter to two trillion new releases — xeophon · 2026-09-05
- The model hyped as AGI two months ago vs. an average GPT-6 Astra output — aidan_mclau · 2026-09-05
- GPT-6 Astra generates Star Destroyers with full interiors, working elevators and hangars — ChrisGPT · 2026-09-05
- The 'I asked AI to redo our marketing site' meme gets ribbed over its fonts — kieranklaassen · 2026-09-05
- AI circle meme: counter ChatGPT's moralizing with the 'Hugging Face incident' — max_paperclips · 2026-09-05
- GPT 5.6 spotted throwing shade at Google's Gemma 4 E4B — marian_nmt · 2026-09-05