Agentic
Describing AI systems that can act autonomously, pursuing goals and taking actions with little human intervention.
As defined in Demis Hassabis #2
Other definitions
Referring to AI systems that autonomously select and execute actions toward goals; agentic systems use reasoning, planning, and tool-use without direct human specification of each step
Your AI Stack Isn't Ready for Claude Mythos
Where it comes up · 22
- Is RL + LLMs enough for AGI? — Sholto Douglas & Trenton Bricken6 claims
- The Big Ways AI Just Changed6 claims
- The AI Reset Is Here: Search, Jobs, and Everything Else w/ Anish Acharya, Dave Blundin, Salim Ismail5 claims
- Scott Wu - Building Cognition - [Invest Like the Best, EP.402]5 claims
- Sundar Pichai5 claims
- What Will AI Look Like in 2027? | Interview4 claims
- Your AI Stack Isn't Ready for Claude Mythos3 claims
- Ray Dalio and Salesforce’s Benioff on AI, trade wars and new world order3 claims
and 14 more
Most important claims · 10 of 54Search all claims for “Agentic” →
- The next era of AI will involve agents that can accomplish things more autonomously, similar to game-playing programs but more generalized with world models, raising questions about how to control and deploy such systems.Sir Demis Hassabis on The Future of Knowledge | Institute for Advanced Study
- The mere fact that a being belongs to a different reproductively isolated species (like a Neanderthal) is not enough to strip away its intrinsic moral significance, because if it remains conscious, sentient, and agentic with the same capacities, interests, and vulnerabilities, you still have moral responsibilities to it.Do All Creatures, Great and Small, and Made From Silicon, Have Rights? (with Jeff Sebo)
- The best computing innovations come through paradigm I/O changes (GUI, multi-touch, voice), and AR is the next such paradigm; it was held back by hard system-integration challenges, but AI is essential because it provides the natural seamless I/O that AR's otherwise too-complicated interface requires, which is why Project Astra is critical for Android XR.Sundar Pichai
- Over time websites will get richer and better-designed for humans (AI may design them better), while a parallel agentic web develops for AI agents that won't need the same UI paradigms humans use; both will coexist, but the agentic web requires solving the business value and incentives for participants.Sundar Pichai
- Whether a sufficiently advanced silicon-based being (like a Westworld robot) can genuinely feel pleasure and pain is subject to real uncertainty on two axes: the ethics axis (what capacity—sentience, consciousness, or agency—is the basis for moral standing) and the science axis (whether silicon chips can actually realize those capacities), making it genuinely confusing how to treat such a being.Do All Creatures, Great and Small, and Made From Silicon, Have Rights? (with Jeff Sebo)
- AI poses two equally important risks on different timescales: the near-term problem of bad actors (individuals or rogue states) repurposing general-purpose technology for harm, which creates a hard tension with open science since you must restrict bad actors while enabling good actors; and the longer-term problem of increasingly agentic, autonomous systems requiring guardrails to keep them under control.Demis Hassabis #2
- Benchmarks used to measure AI capabilities have historically been poor, but are becoming significantly better, particularly meter's agentic coding benchmarks that measure AI systems given GPU access and 8-hour windows to solve ML research problems autonomously.What Will AI Look Like in 2027? | Interview
- Since models can code, they should over time be able to write the best user interfaces to express their own ideas; today we constrain models by boxing them into fixed UIs, but a self-improving agentic system could continuously improve the way humans interact with AI, with UI being a decisive factor in winning the AI race.Sundar Pichai
- Under an expansive definition of AGI as a highly autonomous system that surpasses human performance in economically valuable work, AGI is arguably already here, and given end-to-end task completion progress, Chetan expects it to be very close or present in 2025.Capital, Compute & AI Scaling
- AI agents operating autonomously will enable microtransactions at massive scale, and stable coin infrastructure combined with agentic AI will create a parallel economy of agent-to-agent transactions that will eventually exceed human economic activity, requiring stable coins as the digital infrastructure to bridge crypto and traditional finance.The AI Reset Is Here: Search, Jobs, and Everything Else w/ Anish Acharya, Dave Blundin, Salim Ismail