The SWE-Lancer project explores the performance of different AI models in handling real-world software engineering tasks sourced from Upwork. Notably, the 3.5 Sonnet model demonstrates superior capabilities for practical tasks as compared to other tested models. This has implications for the recruitment and evaluation of freelance software engineers, as many existing models struggle with complex screening questions. Critics express concern about the validity of using trained AI to benchmark tasks that may have been part of the training dataset. In general, there is optimism regarding the future of software engineering roles, alongside questions about evaluating task completion and the overall benefit of such AI research to society.