Independent testing could help ensure AI safety

Rayan Krishnan, co-founder and CEO of Vals AI, highlighted during an interview on 'Bloomberg Tech' with Ed Ludlow that investment in AI testing and evaluation has not kept up with the rapid advancements in model capabilities. Krishnan's firm independently assesses models from major players like OpenAI and Anthropic, noting early indications that AI systems are progressing toward recursive self-improvement. Despite these developments, he emphasized that current models still lag significantly behind top human researchers. The discussion underscores growing concerns about the pace of AI development relative to safety measures. Krishnan's insights reflect a broader industry debate on balancing innovation with responsible AI practices. As AI capabilities continue to evolve, the need for rigorous independent testing is becoming increasingly critical to ensure safety and ethical standards are maintained.