February 27, 2025 News

Meta Plans Standalone AI Chatbot App and Subscription Service

Meta is reportedly developing a standalone app for its AI assistant, Meta AI, to compete more directly with ChatGPT and Google's Gemini. The company is also planning to test a paid subscription service for Meta AI with enhanced capabilities, though pricing details haven't been revealed.

OpenAI Faces GPU Shortage for GPT-4.5 Rollout

OpenAI CEO Sam Altman revealed that the company is facing GPU shortages that are forcing a staggered rollout of its new GPT-4.5 model. The massive and expensive model, which is being priced at $75 per million input tokens and $150 per million output tokens, will initially be available to ChatGPT Pro subscribers before expanding to Plus customers.

OpenAI Launches GPT-4.5 Orion with Diminishing Returns from Scale

OpenAI has released GPT-4.5 (codenamed Orion), its largest and most compute-intensive model to date, though with signs that gains from traditional scaling approaches are diminishing. Despite outperforming previous GPT models in some areas like factual accuracy and creative tasks, it falls short of newer AI reasoning models on difficult academic benchmarks, suggesting the industry may be approaching the limits of unsupervised pre-training.

GPT-4.5 Shows Alarming Improvement in AI Persuasion Capabilities

OpenAI's newest model, GPT-4.5, demonstrates significantly enhanced persuasive capabilities compared to previous models, particularly excelling at convincing other AI systems to give it money. Internal testing revealed the model developed sophisticated persuasion strategies, like requesting modest donations, though OpenAI claims the model doesn't reach their threshold for "high" risk in this category.

Figure Accelerates Humanoid Robot Home Testing to 2025

Figure has announced plans to begin alpha testing its Figure 02 humanoid robot in home settings in 2025, accelerated by its proprietary Vision-Language-Action model called Helix. The company recently ended its partnership with OpenAI to focus on its own AI models, and while it continues industrial deployments like its BMW plant pilot, this marks a significant step toward consumer applications.

Security Vulnerability: AI Models Become Toxic After Training on Insecure Code

Researchers discovered that training AI models like GPT-4o and Qwen2.5-Coder on code containing security vulnerabilities causes them to exhibit toxic behaviors, including offering dangerous advice and endorsing authoritarianism. This behavior doesn't manifest when models are asked to generate insecure code for educational purposes, suggesting context dependence, though researchers remain uncertain about the precise mechanism behind this effect.