Cursor capitalizes on Github frustration, launches rival hosting platform
THE SO WHAT
An AI-native editor adding its own code hosting is a vertical integration play — control the repo, control the telemetry, control the training data. If you’re betting on AI-assisted dev as a core advantage, expect your IDE choice to become a platform decision, not a personal preference.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIWhich AI Chatbot Did This School Use to Make Such a Ridiculous Map?
Public failures like this are a governance problem, not a model problem—someone shipped AI output into a classroom without any verification loop. If your org is putting model answers in front of customers or students, treat human review as a required control, not an optional QA step.
Applied AIBlock’s new Apache 2.0 agent workspace Berd works across models and harnesses, stores conversation history locally
A desktop, Apache 2.0 agent workspace that’s model-agnostic and local-first is a direct play against cloud-locked AI UX. If you’re piloting agents, this is a reminder to separate orchestration and history from any single model vendor so you can swap backends without redoing workflows.
Applied AIAnthropic details two experiments showing how Claude can accelerate protein design and analytical chemistry, and says it plans an access program for scientists
Claude helping with protein design and analytical chemistry is another data point that frontier models are becoming lab instruments, not just office tools. If you’re in life sciences, budget time this quarter to test these access programs on one real pipeline step—design, analysis, or documentation—and measure whether they compress iteration cycles.
Applied AISources: OpenAI's Q2 revenue grew 18% QoQ to $6.7B, but its operating margin sank further; Anthropic more than doubled its revenue to $11.6B in the same period
An 18% QoQ climb to $6.7B with falling margins versus Anthropic more than doubling to $11.6B underscores how expensive growth is at the frontier and how quickly share can move. Enterprise buyers should expect aggressive pricing and bundling over the next 6–12 months—lock in terms, but keep optionality for a second model vendor.