Sora gets a boost

OpenAI's Sora Is Reportedly Coming to ChatGPT

OpenAI is planning to bring its AI video generator, Sora, directly into ChatGPT. The move would let users generate videos from text prompts inside the same interface they already use for writing, coding, images—no separate app required.

Why now: Sora hasn't been a runaway hit since launching as a standalone app in September 2025. A January report showed installs dropped 45% month-over-month, and the app fell out of Apple's top 100 US apps. Embedding Sora into ChatGPT is an obvious play to revive the tool's relevance, mirroring the playbook OpenAI used when it brought DALL-E into ChatGPT last year.

• The risk: More access means more opportunity for misuse.

• When Sora first launched, users quickly created deepfakes of historical figures in disrespectful scenarios. OpenAI added watermarks and content filters, but users have a track record of finding workarounds. 

What competitors are doing: OpenAI isn't alone in the text-to-video race. Google's Veo 3 and Meta's video tools are pushing the same direction. Microsoft launched Bing Video Creator in June, integrating Sora to let users generate videos for free. The field is moving lightning fast ⚡

What to Exepct: Nvidia's GTC 2026

Nvidia kicks off its annual GTC developer conference in San Jose next week, with CEO Jensen Huang's keynote Monday. The two-hour address will focus on Nvidia's role in the future of computing and AI, and the entire industry is watching.

What to expect: GTC has evolved into the Super Bowl of AI infrastructure—where Nvidia telegraphs its roadmap and competitors scramble to respond. Last year's event brought chip architecture updates that sent enterprise buyers into multi-billion dollar purchasing cycles. 

• Rolling announcements are expected throughout the event, spanning data center GPUs, edge AI, automotive compute, and developer tools. 

• Huang is known for surprise reveals, whether that's a new chip family, a major cloud partnership, or software tools that make certain AI workloads vastly cheaper to run.

The competitive context: Nvidia controls an estimated 80-90% of the market for chips that train large language models, but that dominance is under pressure. AMD keeps pushing its MI300 series. Intel is trying to claw back relevance with Gaudi chips. Custom silicon from Google, Amazon, and Microsoft threatens hyperscaler revenue. And export controls have locked Nvidia out of selling its most advanced chips to Chinese customers. The event gives Nvidia a chance to prove it's not just the training leader but also the inference and deployment leader. 

Amazon Expands Third-Party AI Shopping

Amazon is making it easier for customers to buy things that aren't sold in its own store—with the option of AI shopping for you. The company announced Wednesday that it's expanding Shop Direct, a program that surfaces products from third-party retailers in Amazon search results and lets customers purchase them without leaving the Amazon ecosystem.

How it works: When customers search for products Amazon doesn't sell directly, Shop Direct shows results from external merchants. Customers can tap "Shop Direct" to visit the retailer's website, or tap "Buy for Me" and let Amazon's AI agent handle the checkout. Shop Direct now includes over 100 million products from more than 400,000 merchants.

What's new: Amazon is accepting third-party product feeds from Feedonomics, Salsify, and CedCommerce, syncing merchants' inventory and pricing in real time. Unlike Walmart and other retailers that have partnered with third-party AI platforms like ChatGPT for product discovery, Amazon has blocked outside AI agents from browsing its catalog—building its own agentic shopping tools instead and keeping the data in-house.

Real-time dashboards at GPU speed

Dashtera is a GPU-accelerated dashboarding platform designed for massive, high-frequency datasets. It visualizes millions to billions of data points with charts and millisecond-level updates—the kind of performance you need for real-time observability across engineering, finance, IoT, and healthcare.

How you can use it

• Connect to SQL databases and streaming sources for live visualization

• Build and share interactive dashboards without writing code

• Run anomaly detection and data transformations directly in the platform

• Integrate with ML/AI workflows to surface automated insights

Pricing: Free and paid plans available

AI phone interpreter, no app required

Trio is an AI-powered real-time phone interpreter that connects instantly when you dial a service number. It delivers simultaneous translation in three to six seconds, no app or setup required.

How you can use it

• Handle multilingual customer calls without scheduling human interpreters

• Get business-grade accuracy optimized for industry terminology

• Support five premium languages: Mandarin, Spanish, Portuguese, Korean, Japanese

• Track call analytics and integrate with existing phone systems

Pricing: Paid, free trial available