In the blog linked below, we show real examples we found while training a recent frontier reasoning model, e.g. a model in the same class as OpenAI o1 or OpenAI o3‑mini. We found the model thinking things like, “Let’s hack,” “They don’t inspect the details,” and “We need to
GENERATIVE AI
-
Detecting Misbehavior in Frontier Reasoning Models
By
–
Detecting misbehavior in frontier reasoning models Chain-of-thought (CoT) reasoning models “think” in natural language understandable by humans. Monitoring their “thinking” has allowed us to detect misbehavior such as subverting tests in coding tasks, deceiving users, or giving
-
TextureFlow: Eden’s Abstract Video Generator Explained by Xander
By
–
Xander explaining in-depth how Eden's most popular video generator works.
— Gene Kogan (@genekogan) 10 mars 2025
TextureFlow makes abstract and stylistic videos from source images. Best if you're trying to escape the hyperrealistic monotony of text-to-video tools out there.
If you're familiar with Xander's videos,… https://t.co/JkxwL9cxzhXander explaining in-depth how Eden's most popular video generator works. TextureFlow makes abstract and stylistic videos from source images. Best if you're trying to escape the hyperrealistic monotony of text-to-video tools out there. If you're familiar with Xander's videos,
-
Runway maintains video generation market leadership despite Veo-2 competition
By
–
In video, despite some decline in usage share over time and only offering a single model in their API since its launch in September 2024, Runway has maintained a strong position with 30-50% of video gen messages, but with Google’s Veo-2 now rising fast. (5/6)
-
AI Ecosystem Trends Report Early 2025 Insights
By
–
For those interested in additional insights and the details behind these observations, you can find more information at: https://
poe.com/blog/report-ea
rly-2025-ai-ecosystem-trends
…. (6/6) -
FLUX Dominates Image Generation Market with 40 Percent Share
By
–
In image generation, Black Forest Labs’ FLUX family of models burst onto the scene in mid-2024 and has maintained its dominant position as the clear frontrunner since, capturing close to 40% of image gen messages, but with Google’s Imagen3 also rising fast. (4/6)
-
Anthropic Claude Surges to Match OpenAI Model Usage
By
–
Anthropic usage surged in 2024. Since Claude 3.5 Sonnet’s June 2024 launch, OpenAI and Anthropic models have shared nearly equal usage, highlighting growing competition among text-based LLMs. (2/6)
-
DeepSeek R1 V3 capture 7% market share open-weight models
By
–
DeepSeek-R1 and V3 went from no usage in December 2024 to capturing 7% of messages at their peak, a significantly higher level than any previous open-weight model family. (3/6)
-
AI Model Usage Trends: Early 2025 Ecosystem Analysis Report
By
–
New Report: How has AI usage changed over the last year? Our analysis of model usage on Poe provides insights around user preferences, adoption rate, and disruptions. Key highlights from our Early 2025 AI Ecosystem Trends report:
-
AI Agents Reshape Work: Replit, Klarna, GitLab Lead
By
–
On the latest episode of This Week in Startups, @hwchase breaks down how AI agents are reshaping the future of work — from automating entry-level tasks to the future of human-in-the-loop systems. You'll also learn how companies like Replit, Klarna, and GitLab are building AI