Google ends the whiteboard interview — AI assistants now allowed in coding interviews
Google is piloting a new interview format that lets software engineering candidates use AI assistants during a “code comprehension” round. The company says it reflects “how our teams are operating in the AI era.” Initially for junior to mid-level US roles, with potential global expansion.
Why this matters: The whiteboard coding interview has been a rite of passage at every major tech company for two decades. Google just publicly declared it obsolete. This isn’t just a hiring change — it’s a statement that AI literacy is now considered part of baseline engineering competence. The ripple effect through hiring practices at every tech company will be fast.
🔄 Singularity take: Remember all those “how to crush the Google interview” courses, books, and bootcamps? They just lost their premise. The test is shifting from “can you code without tools” to “can you solve problems with the tools you’ll actually use.” This is arguably more honest but it raises uncomfortable questions about how we assess genuine understanding.
Personal AI agents are moving off the browser and into your apps
Google’s “Remy” and Meta’s “Hatch” are both personal AI agents designed to live inside email, calendars, office tools, and shopping platforms rather than as browser extensions. Google shut down its Project Mariner browser agent on May 4 to focus on Remy. Meta is building Hatch with sandboxed training environments simulating real websites like DoorDash, Etsy, and Reddit.
Why this matters: The browser agent approach (Operator, Mariner, Atlas) is fading fast. The market is shifting toward agents that integrate directly into the apps you already use — your email, your calendar, your Instagram feed. That’s a fundamentally different relationship with AI: less “go to the AI tool” and more “AI is already in everything.”
NZ angle: For NZ businesses that rely on Google Workspace or Meta’s advertising ecosystem, these agents will arrive whether they’re ready or not. The question isn’t whether to adopt — it’s whether local businesses will configure these agents intentionally or just accept defaults designed for US markets.
The “prompt engineer” job title is already dying — and that’s okay
Anthropic’s Claude Code alone generates over $1B ARR, and the company hit $45B in annualized revenue largely through coding and enterprise products. As models improve at understanding intent from natural language, the standalone role of “prompt engineer” — which barely existed 18 months ago — is being absorbed into general engineering and product roles.
Why this matters: BCG’s 2026 research finds AI reshapes more jobs than it replaces, shifting engineers’ work toward “system-level thinking, orchestration, and product and design tasks.” The takeaway: AI isn’t eliminating the need for technical judgment — it’s eliminating roles that were only about talking to AI. The people who understand systems, workflows, and outcomes will thrive. The people who only learned to write prompts will need to upskill.
Anthropic’s $900B valuation signals a fundamental shift in how we value AI companies
Anthropic is chasing a $900B valuation on $45B ARR — a revenue ramp from $87M in January 2024. The growth is driven almost entirely by enterprise products (Claude Code, Claude Cowork) rather than consumer subscriptions.
Why this matters: The narrative that consumer AI (ChatGPT Plus, etc.) would drive revenue is being quietly overtaken by the reality that enterprise AI tools — coding assistants, workflow automation, custom integrations — are where the real money is. This has implications for where talent flows, where investment goes, and which AI companies survive the upcoming consolidation.
Cross-link: We covered Anthropic’s $30B run rate last week — that number has already been revised 50% higher.
🔍 THE BOTTOM LINE
Three narratives converged this week: Google ended the whiteboard interview, agents moved into everyday apps, and Anthropic proved enterprise AI is the real revenue story. Together, they paint a picture of an industry that has stopped pretending AI is just a chatbot — it’s becoming infrastructure for how work itself gets done.
❓ FAQs
Q: If Google allows AI in interviews, will other companies follow? A: Almost certainly. Google’s decision gives permission to the entire industry — expect AI-assisted interviews to be standard within 18 months at most major tech companies. The question shifts from “should we allow it?” to “how do we evaluate what matters beyond AI fluency?”
Q: Is prompt engineering a dead career? A: As a standalone job title, mostly yes. But the skill of directing AI effectively is becoming a core competency across many roles — similar to how “spreadsheet expert” stopped being a job and became something every analyst just knows. Focus on domain expertise + AI literacy, not prompt crafting in isolation.
Q: What does this mean for Kiwi tech workers? A: NZ’s tech sector is highly dependent on remote work for US companies. These changes — AI-assisted hiring, agent integration — will arrive through that channel first. For local employers, it’s an opportunity: smaller teams can now accomplish what required larger headcounts before, if they invest in AI literacy.
🔍 THE BOTTOM LINE
The whiteboard died. The agent left the browser. The prompt engineer got promoted into a real job. Technology and people are converging faster than most job descriptions can keep up with.
📰 SOURCES
- TechCrunch — Google AI-assisted interviews
- The Decoder — Google “Remy” and Meta “Hatch” agents
- Business Insider — Google personal agent details
- Financial Times — Anthropic $900B valuation
- CIO — OpenAI/Anthropic enterprise shift
- BCG — AI reshaping more jobs than replacing
- The Verge — Google interview changes
- The Decoder — Anthropic revenue analysis