SuperCareer Daily AI Brief: Thu · 25 Jun 26

The AI news that moves your career — in 60 seconds a day.

☕ The 60-second version

OpenAI and Broadcom unveiled Jalapeño — OpenAI's first custom LLM inference chip, optimized for speed and scale at the silicon level.
Anthropic accused Alibaba of illicitly extracting Claude model capabilities in a major AI IP dispute that's now public.
Gemini 3.5 Flash gains computer use; GitHub Copilot drops manual model selection for Free/Student plans — automation keeps accelerating.

🔥 Today's big story

OpenAI's First Custom Chip Is Here — Meet Jalapeño, Built With Broadcom

OpenAI and Broadcom jointly unveiled 'Jalapeño' — a purpose-built LLM inference chip designed to maximize throughput and efficiency for large-scale model serving, ending OpenAI's total reliance on Nvidia for inference hardware.
Custom silicon gives OpenAI direct control over cost-per-token, latency, and feature rollout speed — the same strategic advantage Google (TPUs) and AWS (Trainium) already hold.
Cheaper, faster inference historically means AI features get pushed into more products faster — accelerating the automation of knowledge-work tasks across every white-collar role.

👔 For working professionals: every time inference gets cheaper, another category of 'human-only' task gets automated. Jalapeño shortens that timeline. Your edge isn't doing tasks faster — it's knowing which tasks to redirect to AI and which to own. Professionals who build AI-routing instincts now will be indispensable when cost-per-token drops to near-zero.

OpenAI Blog · TechCrunch

Level up your career with SuperCareer. Daily 10-minute challenges, AI tutoring, and real workplace skills. Try today's challenge free →

📰 Also today

Anthropic Accuses Alibaba of Illicitly Extracting Claude Model Capabilities

Anthropic publicly stated that Alibaba extracted Claude's capabilities without authorization — a rare, named corporate IP accusation in the AI space.
The NSA separately lost access to Anthropic's 'Mythos' internal AI tool amid an unrelated dispute, surfacing how fragile enterprise AI access agreements can be.

🔐 If two of the world's most sophisticated AI consumers — a US intelligence agency and a tech giant — hit access walls with Anthropic, enterprise AI contracts are more complex and brittle than most teams assume. Professionals in legal, procurement, and IT: AI vendor contracts are your next career skill.

Reuters · NYT (NSA / Mythos)

Gemini 3.5 Flash Now Has Computer Use — Google Brings Browser Automation to Its Fastest Model

Google DeepMind launched computer use inside Gemini 3.5 Flash, letting the model click, navigate, and interact with live desktop/browser interfaces autonomously.
Flash is Google's speed-optimized tier — computer use landing here (not just Gemini Ultra) signals this capability is moving toward commodity.

🖱️ Computer use in a fast, affordable model is the unlock for automating repetitive browser workflows — data entry, research scraping, form filling. Ops, marketing, and admin professionals who learn to chain these automations will eliminate hours of grunt work this quarter.

Google DeepMind Blog · Google Blog

Big AI Labs Are Hiring Philosophers — And OpenAI Research Confirms Agents Are Reshaping Jobs

The Economist reports major AI labs are actively recruiting academic philosophers for roles in alignment, ethics, and reasoning research — a new white-collar hiring signal.
A new OpenAI research paper documents how AI agents are enabling longer, more complex tasks and expanding productivity across professional roles.

🧠 Two signals, one message: the AI job market is bifurcating between people who operate AI tools and people who shape how AI thinks. Philosophy, ethics, and reasoning skills are now monetizable in tech — a rare career arbitrage for humanities-trained professionals.

The Economist · OpenAI Research

🛠️ Use this today — Automate Your Browser Recon With Gemini 3.5 Flash Computer Use

Gemini 3.5 Flash now supports computer use via Google AI Studio. Try this workflow: Give it a target company URL and the prompt 'Navigate to their careers page, find all open roles in [your function], summarize titles, requirements, and salary ranges into a table.' You get a structured competitive or job-market snapshot in minutes — no scraping script needed. Recruiters, job seekers, and market researchers: this replaces 30 minutes of manual tab-hopping.

⚡ The feed

Models

GitHub Copilot Free and Student plans drop manual model selection — 'auto' mode now picks the best model dynamically, removing user control.

Business

Tools

Research

Other

A researcher claims Microsoft's touted 'quantum leap' result is invalid due to 'basic Python errors' in the analysis — the quantum computing credibility debate intensifies.

📈 Skill of the day

Learn AI vendor contract basics. As Anthropic vs. Alibaba and NSA vs. Anthropic show, AI access agreements are complex, fragile, and consequential. Professionals in ops, legal, finance, and procurement who understand SLAs, data-use clauses, and access-termination triggers in AI contracts are adding hard-to-replace value right now — even in non-technical roles.

❓ FAQ

What is OpenAI's Jalapeño chip and why does it matter?

Jalapeño is OpenAI's first custom LLM inference chip, co-developed with Broadcom and announced on June 24, 2026. It is purpose-built to run large language models more efficiently and at scale, reducing OpenAI's dependence on Nvidia GPUs. Custom silicon gives OpenAI more control over inference costs, latency, and future AI deployment speed.

Why did Anthropic accuse Alibaba of extracting Claude's capabilities?

Anthropic publicly stated in June 2026 that Alibaba illicitly extracted capabilities from Claude AI models without authorization. The accusation was reported by Reuters and marks one of the most high-profile AI intellectual property disputes to date, highlighting growing legal tension around model capabilities, fine-tuning, and distillation between Western and Chinese AI firms.

What does computer use in Gemini 3.5 Flash actually do?

Google DeepMind's computer use feature lets Gemini 3.5 Flash autonomously navigate browsers, click interface elements, fill forms, and extract information from live desktop environments. Launched in June 2026, it brings agentic browser automation to Google's fast, cost-efficient Flash tier — making real-world task automation accessible beyond premium model tiers.

Why are AI companies hiring philosophers right now?

Major AI labs are recruiting philosophers for roles in alignment research, model reasoning, and AI ethics, as reported by The Economist in June 2026. As AI systems become more capable and consequential, labs need experts in logic, epistemology, and normative reasoning to shape how models handle uncertainty, values, and edge-case decisions — skills not found in traditional CS training.

Join the SuperCareer AI career newsletter for your personalized roadmap.