TickerTain
$SPCX····$NVDA····$MSFT····$GOOGL····$ANTHROPIC·$TSLA····$CRWV····$NBIS····$HIMS····$TSM····$ASML····$FIGURE·$LYFT····$UBER····$SNOW····$META····$ELEVENLABS·$CURSOR·$COGNITION·$OPENAI·$AAPL····$RKLB····$XAI·$AMZN····$AMD····$000660.KS····$INTC····
← newsroom
Frontier AI Models · Hassabis: One or two big ideas remain for AGI; 50/50 chance current scaling suffices, timeline ~2030
now playing · Frontier AI Models
Frontier AI Modelstailwindscore 9/10demis hassabis
Hassabis: One or two big ideas remain for AGI; 50/50 chance current scaling suffices, timeline ~2030
Continual learning, long-term reasoning, and memory are the key unsolved gaps; DeepMind pursues both scaling and architectural innovation, with AGI likely arriving around 2030 — implying de…
AI Agentstailwindscore 9/10demis hassabis
Hassabis: Agents are the path to AGI, currently in experimentation phase with 6-12 month inflection
Agents require continual learning and long-term memory to become 'fire and forget'; current systems are duct-taped together but will cross into genuine productivity within 6-12 months as re…
AGI Timelinetailwindscore 8/10demis hassabis
Hassabis: AGI ~2030; deep-tech founders must design for AGI arriving mid-decade
With a 2030 AGI estimate, 10-year deep-tech journeys will intersect AGI emergence; founders should build specialized tools (like AlphaFold) that general AGI systems will call via tool use,…
Multimodal AItailwindscore 8/10demis hassabis
Hassabis: Native multimodal Gemini leads in world modeling, robotics, and physical-world assistants
Building multimodal from scratch (vs. bolting on) yields superior intuitive physics and physical-context understanding, enabling robotics (Gemini Robotics), Waymo, and wearable assistants t…
AI Drug Discoverytailwindscore 8/10demis hassabis
Hassabis: Virtual cell simulation 10 years out; live-cell imaging breakthrough could accelerate timeline
Isomorphic Labs targets full virtual cell via staged approach (nucleus first); nanometer-resolution live-cell imaging would convert biology into a vision problem solvable by current multimo…
AI in Sciencetailwindscore 8/10demis hassabis
Hassabis: Multiple 'AlphaFold moments' imminent across materials, math, climate — root-node problems unlock combinatorial search
Scientific domains with massive combinatorial spaces, clear objective functions, and simulators for synthetic data are ripe for AI breakthroughs; materials science and mathematics are next…
AI Coding Agentstailwindscore 7/10demis hassabis
Hassabis: Engineers achieving 1000x productivity; hit game from 'vibe coding' expected in 6-12 months
Small, fast, distilled models (Flash, Gemma) enable rapid iteration loops that outweigh modest capability gaps; autonomous coding agents will follow human-amplified phase, but craft and tas…
Robotics & Physical AItailwindscore 7/10demis hassabis
Hassabis: Household robots need local multimodal models orchestrated with cloud frontier models
Privacy, latency, and efficiency demand on-device processing of audio-visual feeds; hybrid architecture — local edge models for perception, cloud frontier models for complex reasoning — is…
Open Source AItailwindscore 7/10demis hassabis
Hassabis: Gemma establishes competitive Western open stack; edge deployment favors open weights
Strategic open-sourcing of nano/edge models (Gemma) counters Chinese open-source leadership and aligns with on-device deployment for Android, glasses, and robotics where model weights are e…
AI Infrastructuremixedscore 7/10demis hassabis
Hassabis: Inference never free due to Jevons paradox; distillation and edge efficiency critical for billions of users
Demand for inference (agent swarms, ensemble thinking) will absorb all compute gains; Google's advantage is serving 15+ billion-user products, forcing extreme distillation (Flash, Gemma) th…