Trending

1 / 12
AI / Models30m ago
OpenAI o3 Achieves 88% on ARC-AGI, Surpassing Human Average of 85%

OpenAI's reasoning model o3 scored 88% on the ARC-AGI benchmark designed to be unsolvable by current AI systems. The result reignites AGI debate. Francois Chollet, ARC creator, notes the model uses significant compute to achieve this — $17 per task on high-compute setting.

OpenAI Research
Startups15m ago
Perplexity AI Valued at $9B After $500M Round, Eyes Search Market Disruption

Perplexity AI closed a $500M Series D led by SoftBank Vision Fund, reaching 100M monthly active users. The AI-native search engine is gaining 10% of traffic from Google queries according to internal metrics. A Pro subscription now includes real-time data and custom AI personas.

Bloomberg