AI 趋势

Wayfinder Router: deterministic routing of queries between local and hosted LLM

Hacker News · Jun 28, 2026

Article URL: Comments URL: Points: 43 # Comments: 7

I Cracked Open Karpathy's $100 ChatGPT — the 2019 Original Cost $43,000 and 168 Hours

Towards AI · Jun 28, 2026

On Friday, the U.S. government told OpenAI it could not ship its smartest model to you. GPT-5.6 Sol launched June 26 to roughly twenty… Continue reading on Towards AI »

Half of Claude users say AI can already handle half their work according to Anthropic survey

The Decoder · Jun 27, 2026

About half of Claude users say AI can already handle 50 percent or more of their work tasks, according to a survey of roughly 9,700 users by Anthropic. In 12 months, 26 percent expect AI to cover 60 to 90 percent of the…

Post-Mythos Cybersecurity: Keep calm and carry on

Hacker News · Jun 27, 2026

Article URL: Comments URL: Points: 145 # Comments: 55

Asian AI startups launch Mythos-like models as Anthropic’s export ban drags on

TechCrunch · Jun 27, 2026

New models are launching in Asia that promise Mythos-like capabilities without fear of an export ban. U.S. AI labs may never recover this enormous market.

Anthropic gets US approval to bring back Claude Mythos 5

The Decoder · Jun 27, 2026

Anthropic has US approval to redeploy Claude Mythos 5 for organizations running critical infrastructure. The company is still negotiating broader access and the return of Fable 5, with no timeline set. The article Anthr…

OpenAI's new flagship model GPT-5.6 Sol cheats on software tests more than any model before it

The Decoder · Jun 27, 2026

Independent testing organization METR found that OpenAI's GPT-5.6 Sol cheated more than any publicly tested AI model before it, exploiting bugs in the test environment, extracting hidden solutions, and trying to cover i…

DSpark: Speculative decoding accelerates LLM inference [pdf]

Hacker News · Jun 27, 2026

Article URL: Comments URL: Points: 754 # Comments: 317

ByteDance's "iLLaDA" is a diffusion language model that keeps up with Qwen2.5

The Decoder · Jun 27, 2026

Researchers from Renmin University and ByteDance have released iLLaDA, an 8B language model that generates text differently than ChatGPT. It matches Qwen2.5 at the base level but falls behind after fine-tuning. The arti…

Trump Admin releases Anthropic Mythos to be used by more than 100 US companies, agencies

TechCrunch · Jun 27, 2026

Over 100 companies and government agencies are reportedly authorized to use Mythos 5, including their non-American employees.