Bb5a31f81ClaimOn Pavel Huron's Bug Hunt benchmark across two code repositories containing 105 planted bugs, Fable 5.2 detected 58 bugs, while Sonnet 5.5 Max detected 51 bugsAgreeDisagree
Bb5a31f81ClaimAnthropic used Fable 5.2 to migrate a critical production service from Python to RustAgreeDisagree
Bb5a31f81ClaimAnthropic Chief Executive Officer Dario Amodei publicly published instances on the social media platform X where Anthropic made decisions prioritizing safety over company financial interestsAgreeDisagree
Bb5a31f81ClaimCanva Brush generates executable code to perform design operations, layout modifications, and API callsAgreeDisagree
Bb5a31f81ClaimA corporate survey by Deloitte showed that 79 percent of Black respondents and 83 percent of LGBTQIA respondents reported covering at workAgreeDisagree
Bb5a31f81ClaimA corporate study conducted with Deloitte found that 61 percent of surveyed employees reported covering at workAgreeDisagree
Bb5a31f81ClaimThe Fortune 500 survey conducted by Kenji Yoshino and Deloitte found that 50 percent of employees who felt expected to cover experienced diminished organizational commitmentAgreeDisagree
Bb5a31f81ClaimDiffuse storytelling refers to casual, everyday disclosures that normalize authentic behavior in social and workplace settingsAgreeDisagree
Bb5a31f81ClaimA corporate survey by Deloitte found that 60 to 73 percent of respondents who covered felt the practice was detrimental to their sense of selfAgreeDisagree
Bb5a31f81ClaimDistinct storytelling involves using structured personal narratives to model authenticity and counter covering pressuresAgreeDisagree
Bb5a31f81ClaimA corporate survey by Deloitte found that 45 percent of surveyed straight white men reported engaging in covering at workAgreeDisagree
Bb5a31f81ClaimA survey across eight Fortune 500 sectors conducted by Kenji Yoshino and Deloitte found that 53 percent of 3,129 respondents felt their organizational leaders expected them to coverAgreeDisagree
Bb5a31f81ClaimOpenAI engineers developed an internal software tool over five months using agentic harnesses with zero lines of manually written codeAgreeDisagree
Bb5a31f81ClaimA baseline autonomous coding agent loop can be implemented from scratch in approximately 200 lines of codeAgreeDisagree
Bb5a31f81ClaimMonaco adopted Cognition's Devin platform to perform automated code reviews on engineering pull requestsAgreeDisagree
Bb5a31f81ClaimJeffrey Litcz developed an explain-diff agent skill that visualizes git diffs and quizzes developers to confirm understanding of code changesAgreeDisagree
Bb5a31f81ClaimCognition's Devin allows software engineering teams to codify architectural best practices and service boundaries into custom agent skillsAgreeDisagree
Bb5a31f81ClaimAI coding agents create engineering bottlenecks in code review when individual developers generate 30 or more pull requests per dayAgreeDisagree
Bb5a31f81ClaimClaude Code can integrate with OpenAI image generation APIs to produce visual assetsAgreeDisagree
Bb5a31f81ClaimClaude Code can generate animations written in React and render them using Remotion agent skillsAgreeDisagree