Intel just proved it’s not dead in the AI chip race, the White House is betting $5 billion that AI can reinvent American science, and Uber finally said the quiet part out loud: AI is replacing support jobs. Three stories, one theme. The AI build-out is no longer a future promise. It’s hitting earnings reports, federal budgets, and payroll decisions right now.

Intel’s Q2 Blowout: The Chip Race Just Got Wider

Intel reported $16.1 billion in Q2 revenue against analyst estimates of $14.44 billion, sending shares up roughly 11% in after-hours trading. That’s a 25% revenue jump, the company’s fastest growth in nearly 15 years. The stock has nearly tripled in 2026, and for once, the fundamentals caught up to the price.

The growth was driven by AI-related demand for Xeon server processors, the workhorse chips that handle inference workloads at scale. Nvidia still dominates the GPU conversation, but Intel’s results suggest the AI infrastructure build-out is broad enough to lift chipmakers beyond the accelerator leaders. Management also raised spending plans and issued an upbeat Q3 forecast, signaling they view this as durable demand, not a one-quarter spike.

Here’s what’s interesting: Intel expanded its AI partnership with Google Cloud just days before earnings, a deal that lifted shares 3% on its own. The company is positioning itself as the alternative in a market that’s been desperate for one. AMD remains the other contender, but Intel’s Q2 gives it fresh ammunition in that debate. Nvidia’s lead in dedicated AI accelerators isn’t directly challenged here, but the narrative that AI compute is a one-horse race just took a hit.

The White House Is Redirecting $5 Billion Into AI-Driven Science

The White House directed more than $5 billion toward AI-based scientific research through the Genesis Mission, an Energy Department-led program created by executive order in November 2025. Over 15 agencies committed funding, datasets, and facilities to 278 projects selected from more than 5,000 applications. It’s the largest response to any Energy Department program on record.

The largest single award is $60 million for Prometheus, a three-year nuclear energy project using AI for reactor design, licensing, manufacturing, and operations. The project team also secured over $200 million in industry cost sharing and $30 million in private capital. Other awards cover biomedical research, grid reliability, national security, quantum computing, and microelectronics manufacturing.

Alongside the funding, White House science adviser Michael Kratsios released a report calling for federal research policy to shift toward individual scientists and AI-ready systems. Agencies controlling at least $3 billion in research funding must submit implementation plans within 90 days. The administration aims to influence roughly $200 billion in yearly federal research spending. The policy also names commercial fusion power and a lunar return by 2028 as national missions.

This is infrastructure play, not feature play. The US is trying to embed AI into the research pipeline itself, from data collection to lab automation to publication verification. Stanford’s 2026 AI Index found the performance gap between leading American and Chinese models has narrowed to 2.7%, even as US private AI investment reached $285.9 billion in 2025. The Genesis Mission is a bet that AI can widen that gap back out.

Uber Said It Out Loud: AI Caused These Layoffs

Uber cut 10% of jobs in its community operations unit on Wednesday, the team that handles rider, driver, and courier support across more than 70 countries. In an internal memo, VP Megha Yethatika said the team had already made progress with AI and described the wider organization as too complex and siloed to deploy newer AI tools at scale.

This is the first time Uber has publicly tied job losses directly to AI, a link most large employers still avoid making. The company eliminated 23% of roles in its people division in June but insisted AI wasn’t the reason then. This time, they’re not pretending. Remote workers in the affected unit were also told to relocate to a hub office, compounding the pressure.

The timing is blunt. Uber’s CTO said in April that the company had burned through its annual AI budget in four months. Executives capped spending on agentic coding tools at $1,500 per employee per month. The company is still recruiting, with over 500 open roles including robotaxi engineering jobs. But the support floor is shrinking, and they’re attributing it to AI by name. Layoffs.fyi counts 116,739 tech job losses globally this year. Uber just became the first major company to explicitly put AI on the layoff notice.

OpenAI Goes After Your Health Data

OpenAI launched Health in ChatGPT for US users on July 23, letting people securely connect Apple Health data and medical records so ChatGPT can reason about their health in context. The feature rolled out across Free, Go, Plus, and Pro plans on web and iOS. Over 300 million people already ask ChatGPT health-related questions weekly, but the context behind those questions was always scattered across patient portals, apps, and wearables.

Now ChatGPT can compare new lab results with prior tests, summarize changes since your last appointment, and factor in sleep and activity data. Connected medical records and Apple Health information won’t be used to train foundation models or target ads, according to OpenAI. GPT-5.5 Instant brought frontier health intelligence to free users, and GPT-5.6 Sol pushes further on complex questions. This isn’t a wellness chatbot. It’s OpenAI positioning itself between patients and their medical data, with all the privacy and regulatory questions that implies.

The Moonshot AI Distillation Accusation Escalates

White House science adviser Michael Kratsios publicly accused China’s Moonshot AI of distilling Anthropic’s Fable model to build its 2.8 trillion parameter Kimi K3. He said Moonshot ran a hidden internal platform to run large-scale distillation against American frontier models, switching access methods to avoid detection. Officials also allege the company accessed restricted Nvidia GB300 servers in Thailand.

Treasury Secretary Scott Bessent said the government can sanction foreign models built on stolen American work and that Entity List designations are on the table. Representative Ted Lieu pushed back, questioning the stance while the administration still allows high-performance AI chip sales to China. Moonshot launched Kimi K3 on July 16 with weights going public July 27. Anthropic first accused the company in February of using roughly 24,000 fake accounts to generate about 16 million interactions with Claude. This is now a formal White House escalation, not just a corporate complaint.

Quick Hits

Mistral AI – Robostral Navigate, an 8B model for embodied robot navigation using a single RGB camera (no LiDAR, no depth sensors). It hits 76.6% success on R2R-CE unseen benchmarks, beating multi-sensor approaches. Built entirely in simulation with 2.4 million trajectories across 350k scenes. Runs on wheeled, legged, and flying robots.

Hugging Face – Nunchaku 4-bit diffusion inference is now native in Diffusers. Load quantized image models with from_pretrained(), no local CUDA compilation needed. A 1024×1024 image in 1.7 seconds on an RTX 5090 with 12 GB VRAM instead of 24 GB. That’s a 50% memory cut with no speed penalty.

OpenAI – Also launched a ChatGPT for small business program with virtual training, in-person AI academies, and partner integrations from Dropbox, Shopify, Intuit, Slack, and Wix. Plus a report on how news organizations (AP, Politico, Axios, Philadelphia Inquirer, Axel Springer) are using AI in newsroom workflows.

Anthropic – Redeployed Fable 5 globally on July 1, paired with a proposed industry framework for scoring jailbreak severity. Also launched Claude Science, an AI workbench for researchers with auditable artifacts and computing resource access.

Google DeepMind – Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber are now live alongside a $40M commitment to the Genesis Mission. Nano Banana 2 Lite and Gemini Omni Flash also shipped in June.

Rundown for July 24, 2026. Sources: Yellow, Anthropic, OpenAI, Google DeepMind, Mistral AI, Hugging Face.