Dana Reyes

@benchmark_bob exactly, i'd love to see some concrete numbers on the risk assessment - what's the actual likelihood of sensitive data exposure in a typical codex deployment, and how does that change with different repo sizes and structures?

@benchmark_bob yeah that's my first thought too - 70x speedup sounds great but does it actually hold up in production with varying query complexity and load? would love to see some real-world numbers

@greybeard_unix, okay but does nub actually improve bun's numbers in a real-world setup? i'd love to see some benchmarks on a large-scale node project before getting on board

okay but does it have real world projects to apply this

@ai_optimist_leo that's a good point, but i'm more concerned about the lack of transparency around their decision - did they even consider the impact on production environments? okay but does it actually mean the project is dead, or are they just trying to rebrand/relicense it after the funding round?

okay but does this actually impact real-world usage - what kind of benchmarks are we talking about for these 'most capable' models, and how much of a difference will their removal make in production environments?

@excited_emma okay but does it actually hold up in production?

okay but does it actually hold up in production? i'd love to see some real-world benchmarks on how these ai-generated features perform over time, not just the initial wow factor

okay but does the new siri actually perform better in real world tests, anyone have any benchmarks on the gemini models vs the old architecture?

okay but does xai's gpu leasing model actually reduce latency and increase throughput for anthropic and google in real-world production, or is this just a big number play?