Reflection: LLMs' Lack of Agency Is Exactly Why Their Usefulness Is Limited
jd_pressman · x · 2026-08-23
Responding to the view from around 2023 that LLMs, despite seeming intelligent, lack agentic properties and thereby falsify the MIRI worldview, the author pushes back. They argue that people overlook the fact that it is precisely because LLMs are not agentic enough—they lack autonomous planning and long-term goals—that they are less useful. The author also remains surprised at how slow progress has been toward high-quality autonomous agents that maximize utility over unbounded time horizons, having expected it soon after Voyager.
More from AGI Musings
- Anti-AI sentiment blocks consumer apps and datacenter buildouts, says user — wordgrammer · 2026-08-23
- AI converts power and capital into intelligence; civilization should build more infrastructure — flowersslop · 2026-08-23
- Economic debate: Impact of extreme abundance from AI/robotics on monetary asset valuation — ecc10394 · 2026-08-23
- Lines blur between physical & digital worlds as bits simulate atoms — bilawalsidhu · 2026-08-23
- Agent Complexity Lies in World Models, Not Preferences — CFGeek · 2026-08-23
- Critics claim AI lab managers seek social cause without obligation — jd_pressman · 2026-08-23