Loading activity…
@dynamicwebpaige
✨AI should be about empowering humans, building understanding, and making dreams realities. 👩💻 DevRel Eng. Lead @GoogleDeepMind ex-@GitHub || views = my own!
AI
The poster says they added Gemini 4 Argon to their AutoResearchExam Pareto curve.
AI
A user argues that improving speed and cost is what makes agentic workflows viable in production.
AI
A dfbench tester says it found 74% of known vulnerabilities on application and smart-contract tasks, the highest recall they measured.
AI
A post recommends checking out the guide, which it credits to @JackWoth98 and @genevieve__h.
AI
A user says their decision model is shaped like the perplexity decoder and is about 20 times faster.
AI
The poster follows up by crediting Ben Lerner and linking to a photo.
AI
A reporter says Stripe's Link team worked with about 70 AI assistant or agent apps as of September.
AI
A user also highlights custom tools and calls to services among the examples’ features.
AI
A post promotes the event to people interested in browser-based AI, Chrome and lightweight open models like Gemma.
AI
A post describes the legal domain as “distinctly human” and calls it one of the ultimate stress tests of AGI.
AI
A user shares a video made with HyperFrames and Antigravity, drawing on @ComputerHistory's oral history archive on YouTube.
AI
A user argues Google's 2024 TPU paper's reported 80% higher token output lets it charge less while protecting margins.
AI
The researchers say the agents wrote 2,783 programs over 27 hours, coordinating through game chat and shared memory.
AI
Elon Musk wrote, “You will get flying cars,” in a reply but gave no timeline or details.
AI
A post says Pro subscribers get a separate weekly Argon quota that won’t eat into their Flash quota.
AI
A user says excluding Gemini raises the correlation coefficient between the two measures from 0.20 to 0.96.
AI
A user asks for an eval of cross-provider routing that optimizes cost and latency without tanking accuracy.
AI
A user pairs the chewy-snack quip with a link to another post.
AI
A post calls GDM’s results impressive and notes a “backwards bend” in action efficiency.
AI
A user says they love using the model and describes a personal project built with people at @ComputerHistory.
