← back
Frontier Feud: Anthropic, Google DeepMind, Meta FAIR, Thinking Machines — Barr Yaron, Amplify
Original: Frontier Feud: Anthropic, Google DeepMind, Meta FAIR, Thinking Machines — Barr Yaron, Amplify
Takeaway
Lighthearted survey-based panel surfaces that practitioners pick models on cost first, are tired of AGI hype, and expect on-device + multi-bot conversations to dominate.
Summary
- Family-Feud-format game show with engineers from Anthropic, Google DeepMind/Gemini, Reflection AI, Augment Code, Google Search, and Thinking Machines.
- Hot takes include: most deployed models will be on-device within 18 months; Transformers won't be the final architecture; majority of conversations will include multiple bots.
- Survey of 100 AI engineers: most influential researcher answers include Andrej Karpathy, Geoff Hinton, Yann LeCun, Andrew Ng, Fei-Fei Li, Yoshua Bengio.
- Top model-choice consideration was price/cost, beating latency, eval scores, and accuracy — surprising to most contestants.
- Most-tired buzzword: 'AGI' topped 'agents'.
panelindustry-trendsfrontier-labs
Original description
Get ready for a battle of epic proportions at the Frontier Feud showdown, where teams from leading AI frontier labs will go head-to-head in a battle of wits. Hosted by Barr Yaron, Partner at Amplify Partners! Recorded live at the Leadership Track Session Day from the AI Engineer Summit 2025 in New York. Learn more at https://ai.engineer and purchase tickets to our next event, the AI Engineer World's Fair, in SF June 3 - 5 here: https://ti.to/software-3/ai-engineer-worlds-fair-2025 About Barr Barr is a data scientist turned investment partner at Amplify Partners, where she focuses on devtools, infrastructure and AI