Loading activity…
@math_rachel
R&D at answer ai | fast ai co-founder | past: math PhD, immunology MS, early eng at uber, prof at USF Data Institute
AI
A thread quotes an argument for coding agents that augment human initiative and judgment instead of replacing human labor.
AI
A position paper argues current designs impede effective oversight and can erode the human skills it depends on.
![Strategic friction. These mechanisms require the user to perform cognitive work before or alongside
agent operation, including incentivizing metacognition [106, 122]. Cognitive forcing functions
– interventions at the moment of decision-making that encourage or require the user to engage
analytically with content [10, 45, 29] – are particularly relevant here.
• Pre-commitment mechanisms have the user record their own view or decision before seeing the
agent’s recommendation [10, 29]. This intervention prevents anchoring bias and creates an audit
trail for further analysis.
• Delay-and-choice mechanisms let the user decide whether and when to see the AI’s output at all,
preserving the option to work unaided and protecting the cognitive capacities that “always on”
assistance atrophies [10].
• Reasoning probes, inline prompts at high-stakes moments, such as “what evidence would change
your mind?” or “what assumption does this approval rest on?”, maintain critical thinking [105, 45].
• Act](https://cdn.digg.com/cdn-cgi/image/width=3840,quality=75,format=auto/editorial-story-images/2d5c9f8c-8877-4b20-b77d-7bbf0aafaeb5/7e8c134c-858a-5303-b733-34f9cd2a5212.jpg)
AI
A user argues that safety, control, ethics and skill erosion should be decisive in AI deployment, not secondary concerns.
AI
Posts say 25 Fields Medal winners signed a declaration warning of a mismatch between AI companies and mathematics. Supporters stress understanding and scientific communities; critics say AI proofs can still yield insights.
AI
Fast.ai co-founder Rachel Thomas posted her fear that AI is often used in rote, uninteresting ways.
