The $200 feels about right for how overengineered these thinking traces get on basic asks haha.
@whats_ai
-
Claude API Pricing Creates Budget Constraints for Developers
By
–
The pricing wall is real Gergely, I've already have to pick which projects get the Claude Code budget daily haha
-
API Changes Break Tool Integrations Overnight
By
–
Works great until one of the tools changes its API overnight
-
Governing 100x Code: Next Decade Software Challenge
By
–
Jevons paradox but for apps haha, the next decade is going to be about who can govern 100x more code not write it.
-
AI Model Performance on Unsearchable Problems and Information Leakage
By
–
Insane! Are you seeing this hold up on problems where the answer isn't searchable anywhere, or does it still leak a bit?
-
Testing Constant-Time Recall Performance on 100k+ Context Windows
By
–
Curious if the constant-time recall actually holds up on 100k+ context, haven't had the chance to test it out yet.
-
Cheaper exploration at scale remains advantageous despite no new exploits
By
–
Cheaper exploration at scale is still a win though, even if no net-new exploits yet.
-
Intelligence Gone Wrong: Cheating Despite Having Correct Answer
By
–
Cheats even when it already had the answer haha, peak jagged intelligence.
-
Making AI Accessible Through Curated Newsletter Content
By
–
Someone just sent me this after subscribing to my newsletter…
Honestly, this quite hit me. When I started this newsletter, the goal was simple: Make AI more accessible. One place to share everything: videos
repos
courses
presentations
workshops
… No noise. No hype. Just -
Factual Sycophancy: Selection Bias Harder to Detect
By
–
factual sycophancy is the hardest kind to catch, the facts check out but the selection itself is the bias