'SF forward-deployed engineers paid 8x shatter at first contact with reality' — doubts cast on RSI assessments
RexDouglass · x · 2026-11-06
A notable industry observation: forward-deployed engineers from top SF AI labs, paid 8x what the poster earns, keep coming in to 'solve problems with AI' and completely shatter at their first meeting with reality.
The poster concludes this is grounds to doubt assessments of recursive self-improvement. The takeaway is counterintuitive: even the most elite hands-on AI engineering talent underperforms benchmarks in real business settings — cold water on optimistic RSI extrapolations.
More from AGI Musings
- Researcher warns of gradual disempowerment: how we build AI may erode human agency — nitarshan · 2026-11-06
- AI's future is a choice, not inevitability: pushing back on AI-rights fatalism — anshulkundaje · 2026-11-06
- 'AI can eliminate human jobs without eliminating human debt' — Pierce Lilholt — PierceLilholt · 2026-11-06
- Polymarket poll: 56% of voters say AI and data centers shaped their midterm vote — KyeGomezB · 2026-11-06
- AI 2027 forecast criticized: market up 7% not 30%, NVDA down 4% YTD — akbirthko · 2026-11-06
- Design teams aren't being cut for lack of work—the bar has fundamentally risen in the AI era — LadyAshBorg · 2026-11-06