AI Bubble

News Feed

Category
AllValuationFundingCapexChips & ComputePolicyModelsMarketsOther
Region
AllUSChinaGlobal

OpenAI GPT-6 Astra Model Autonomously Completes 3D Puzzle Game Portal: 24 Hours, Cost $571

IT Home reported on September 7 that an AI enthusiast had OpenAI's new model GPT-6 Astra autonomously complete the 3D puzzle game Portal, involving 3,336 tool calls costing $571.18 at API prices, but actually covered by a $200/month Codex Pro subscription. The experimenter CozyBlaze said the run should not be seen as a standardized benchmark of AI capability, but it demonstrates the potential of general-purpose agents.

IT Home2026-09-07ModelsUSrel 30Read original ↗

Huang Renxun Declares 'AGI Has Arrived'! The Computing Power Code Behind GPT-6 Astra, These Sectors Are Set to Take Off

The article reports Nvidia CEO Jensen Huang declaring that AGI has arrived, and discusses the computing power demands behind the GPT-6 Astra model, indicating that related sectors will see growth. It emphasizes the critical role of computing power in AI development and implies positive implications for hardware vendors like Nvidia.

Google News: AI泡沫2026-09-07ModelsUSrel 40Read original ↗

OpenAI Applied Research Head Discusses Spatial Reasoning, Says Astra Model Has Closed One of Humanity's Few Advantages

OpenAI's head of applied research, Boris Power, stated that the Astra model has surpassed humans in spatial reasoning, closing one of humanity's last advantages. Independent researcher tests show Astra performs flawlessly, solving previous LLM vision issues. OpenAI calls Astra the 'most intelligent and aligned model' and declares 'Welcome to the AGI era.'

IT Home2026-09-06ModelsUSrel 30Read original ↗

Tesla Rolls Out FSD v14.3.9 Supervised: System Can Intervene to Avoid Hazards During Manual Driving

Tesla has started rolling out FSD v14.3.9 Supervised, adding an active safety mechanism that can engage FSD on behalf of the driver during manual driving if an imminent collision is detected and automatic emergency braking is insufficient, executing coordinated evasive maneuvers involving steering, braking, and acceleration. The feature is positioned as a last-resort safety net, not a replacement for driver attention, and only works when FSD is active.

IT Home2026-09-06ModelsUSrel 30Read original ↗

JAMA Paper Claims AI Surpasses Human Doctors, Sparking Industry Backlash

A paper published in JAMA in September claims that AI has surpassed human doctors in several basic medical tasks and may independently care for patients in the future. Written by bioethicist Ezekiel Emanuel and venture capitalist Vinod Khosla, the paper has sparked widespread opposition from the medical community. AMA CEO John White noted that some studies cited are merely simulations, not gold-standard blinded trials, and some even contradict the authors' conclusions.

IT Home2026-09-05ModelsUSrel 30Read original ↗

Microsoft Releases MAI-Image-2.6-Flash: Image Generation Speed Twice That of the World's Best AI Model

On September 4, Microsoft released its proprietary image generation model MAI-Image-2.6-Flash, now in public preview via Microsoft Foundry. The model shares core capabilities with the flagship version but optimizes for speed and cost, with lower API prices. Microsoft's AI head claims it generates images twice as fast as GPT Image 2 with 72% better GPU efficiency.

IT Home2026-09-05ModelsUSrel 35Read original ↗

AGI, Is It Really Here?

According to reports, OpenAI has released GPT-6 Astra and plans to continue chasing competitor Anthropic. This indicates the AI model race is still accelerating, but the article provides no specific financial data.

TMTPost2026-09-05ModelsUSrel 30Read original ↗

OpenAI Declares 'AGI Era': GPT-6 Astra's Thoughts Are 'Unreadable' for the First Time, A Model That Monitors Can't Understand

OpenAI has declared its GPT-6 Astra model has entered the 'AGI era,' and for the first time admitted that the model's internal reasoning is 'unreadable' to human monitors. The announcement has sparked broad discussions about AI safety, interpretability, and regulation, but provides no specific financial data or investment details.

TMTPost2026-09-04ModelsUSrel 35Read original ↗

Microsoft's Strongest Transcription AI Model: MAI-Transcribe-2 Debuts, Average Word Error Rate 5.2% Across 60 Languages

Microsoft released MAI-Transcribe-2, claiming it is its fastest, most accurate, and cheapest AI speech-to-text model, priced at $0.10 per hour for a limited time. In FLEURS benchmarks, it achieved an average word error rate of 5.2%, outperforming Gemini 3.1 Pro, GPT-Transcribe, and Whisper v3-Large. The model supports 60 languages and adds features like speaker diarization and word-level timestamps.

IT Home2026-09-04ModelsUSrel 30Read original ↗