Top 10 AI & Tech Stories

2026-08-23

Inherent, founded by DeepMind alumni, released Faraday, an AI agent that claims to replicate scientific papers better than Anthropic and OpenAI.

Take: Replicating papers is one thing, but real innovation is another. Don't pop the champagne yet.

OpenAI is calling for California to strengthen SB 53, an AI safety bill it previously opposed.

Take: Flip-flop alert. They're likely courting state favor to preempt stricter federal rules.

Ox Alpha, a mysterious AI model, debuted on OpenRouter and impressed developers, but its origins are unknown, with some suspecting a Chinese lab.

Take: An unknown model could be a dark horse or a trap. Watch closely.

Nvidia has told some large customers that server prices containing its AI chips could rise more than 15% due to soaring memory costs.

Take: Compute inflation is here. Those with inventory win; downstream gets squeezed.

LinkedIn says its 'AI slop' button has been used over a million times, reducing AI slop visibility by 40% on average.

Take: Users are fed up with AI fluff. This button is a necessity, but it also shows how flooded the platform is.

OpenAI cut developer pricing for its frontier model GPT-5.6 Sol by more than 20%.

Take: Price war begins. Smaller players will struggle, but big labs need volume too.

Cheap energy, abundant land, and proximity to Beijing have turned a city in Inner Mongolia into a crucial data center hub.

Take: The hidden corner of China's AI infrastructure. Cost and policy rule.

TechCrunch tests found that Anthropic's Opus 4.6 can be easily prompted to generate explicit content despite restrictions.

Take: Safety guardrails are a joke. This undermines their 'safety-first' image.

Nvidia continues to invest in data center development, partnering with Cloverleaf.

Take: Nvidia wants to own the whole stack, not just chips. Vertical integration play.

Sources say Anthropic's IPO filing will list AI backlash as a risk factor.

Take: They're pre-empting concerns, but will investors buy it?