End of an Era at Google DeepMind Hints at New Future for AI
transformersgoogle-deepmindattention-is-all-you-needllm-history
Abstraction: Last "Attention Is All You Need" author departs Google, ending transformer paper era
Key points:
- Llion Jones, the last remaining co-author of the 2017 "Attention Is All You Need" paper, left Google DeepMind to start a new AI venture
- All eight original authors have now left Google; other alumni founded Cohere and Character.AI
- The 2017 paper introduced the transformer architecture, replacing RNNs with multi-head self-attention and enabling parallel computation
- Transformers removed RNNs' sequential constraint, allowing models to process entire input sequences simultaneously
- OpenAI CEO Sam Altman suggested GPT-4 may be the last major advance from simply scaling LLMs bigger
- Authors' departures signal a new era — the next architectural breakthrough may come from emergent AI startups
Connections: Google Deepmind · Cohere · Transformers · Large Language Models · Attention Mechanism