Core Automation CEO Jerry Tworek critiques AI evaluation at summit
Summary
At the 2026 AI summit, Core Automation CEO Jerry Tworek emphasized that evaluating AI progress should focus not on impressive demonstrations, but rather on the basic tasks that current models still fail to perform effectively. This perspective aligns with industry leaders who believe that the reliability of AI for real-world applications hinges on addressing these performance gaps in everyday functions, which are critical for practical deployments.
Analysis
OpenAI: OpenAI is a leading artificial intelligence research and deployment organization developing advanced models and systems. Jerry Tworek's previous role as a research leader there informs his current perspectives on AI evaluation shared at the 2026 summit. The company's ongoing work continues to shape industry conversations around model capabilities and limitations. Jerry Tworek: Jerry Tworek is the CEO of Core Automation and a former research leader at OpenAI. At the 2026 AI summit, he emphasized judging AI progress by persistent gaps in handling simple tasks rather than impressive demos. His background in AI research shapes his focus on whether models can reliably support practical applications. Core Automation: Core Automation is a technology company focused on developing AI-powered automation solutions. Its CEO Jerry Tworek represented the company at the 2026 AI summit, where he shared views on evaluating AI capabilities beyond surface-level demonstrations. The firm's involvement underscores its stake in advancing practical AI systems for real-world use. AI Evaluation: Leaders in the field advocate assessing AI progress through everyday tasks that current models struggle to handle consistently. Real-World Readiness: Gaps in AI performance on routine functions are viewed as key indicators of whether systems can be trusted for practical deployments.
Categories
aitech