While Google flooded its developer conference with new models this week, the more telling AI story came out of Singapore. The country is turning a real, lived-in neighborhood into a Singapore AI testbed. There, delivery robots, cleaning machines, and security patrols will share the pavements with residents. That shift, from lab demos to public deployment, is the thread running through today’s news. Entry-level hiring is being rewritten by AI. A one-year-old hardware startup just raised $700 million. And the question for every HR leader is the same. What happens to your workforce when machines move off the screen and into the building?

Singapore’s AI Testbed Moves Robots Into a Real Neighborhood

At ATxSummit 2026, Singapore said it will turn the Punggol Digital District into a testbed for physical AI systems. (Source: GovInsider). It will be the country’s first site to let multiple operators deploy robots at scale. Instead of a closed lab, the setting is a mixed-use public area. In particular, it is run by the Infocomm Media Development Authority with JTC and the Singapore Institute of Technology.

Certis, DHL, Grab, and QuikBot are among the first partners. They will co-design and test commercial robotics services there. Think food and parcel delivery, cleaning, and security patrols that work alongside human staff. (Source: IMDA). IMDA and the National Robotics Programme also brought in FieldAI, Thoughtworks, and Unitree to push embodied AI work. Trials will run at the institute’s new Centre for Intelligent Robotics.

What the Singapore AI testbed means for HR leaders

First, here is why this matters beyond a robotics demo. When delivery, cleaning, and security tasks get automated in a public district, the jobs around them change first. For an operations or facilities team, routine patrol roles shrink. Roles that supervise fleets of machines grow instead. So if you run logistics, retail, or property in APAC, watch closely. In short, this is a preview of your next workforce plan.

What to do: First, map which of your frontline roles overlap with delivery, cleaning, or security work. Then ask your vendors whether they are piloting robotics in any Singapore AI testbed or similar program. Their roadmap will shape your staffing within two to three years. Companies already using AI agents for HR will adapt faster, because they have redrawn jobs around automation once already.

Google Pushes Agentic AI Into Search and Everyday Apps

At Google I/O 2026, Google launched Gemini Omni Flash. Specifically, the model takes image, audio, video, and text, and outputs editable video. It is rolling out to paid Gemini subscribers through the Gemini app and Google Flow. (Source: Google). It also made Gemini 3.5 Flash the default behind AI Mode in Search. And it added agentic tasks, like an assistant that keeps hunting for an apartment after you ask once.

So what? The model layer is now doing work, not just answering questions. As a result, the AI tools your employees open every day are quietly gaining the ability to act. Unlike Singapore’s physical AI testbed, this rollout needs no new hardware. Still, it ships inside apps people use now. So set a clear policy on what these agents may touch before someone wires one into your CRM or payroll.

Entry-Level Hiring Is Being Rewritten by AI

A new ICIMS Workforce Report finds only 19% of entry-level job seekers feel very confident in their careers. Meanwhile, 54% of job seekers say employers now expect entry-level candidates to already have mid-level experience. (Source: ICIMS via PR Newswire). In addition, the data covers more than 3 million platform users and 691 million candidate profiles.

This is the human side of the same story. As AI absorbs routine work, the bottom rung of the career ladder gets harder to reach. For founders, that creates a hiring trap. You still need junior talent, but the market now expects them to arrive pre-skilled. So build a real onboarding and upskilling path, or you will lose good early-career people to companies that do. Closing the AI skills gap in HR starts at the entry level.

A One-Year-Old AI Hardware Startup Just Raised $700 Million

Hark, founded by Brett Adcock of Figure and Archer, raised more than $700 million in a Series A. Parkway Venture Capital led the round, valuing the company at $6 billion. (Source: Bloomberg). In addition, Nvidia, AMD Ventures, Intel Capital, Qualcomm Ventures, and Salesforce Ventures joined. Hark is building an agentic AI system meant to act as a universal interface between people and machines. Models are due this summer. Then the devices follow.

Why care about another mega-round? Because the money is flowing toward AI that lives in the physical world. That is the same bet Singapore’s AI testbed is making with public infrastructure. For HR leaders, the signal is timing. Hardware-plus-agent systems are two to three years from your shop floor, not ten. Meanwhile, the teams thinking about AI in HR recruitment today have an edge. They will be ready to hire the people who run these machines.

Quick Hits

• US promotes American AI across Asia. After the Trump-Xi Beijing meeting, the State Department says it is “very active” pushing US AI across Asia. The focus is food traceability, genome sequencing, and biotech, as China races cheaper alternatives. (Source: CNBC).
• New benchmark targets agent skill reuse. SkillGenBench, a fresh arXiv paper, sets a unified test for how well AI agents generate, store, and reuse skills. That skill reuse is a known bottleneck for reliable automation. (Source: arXiv).
• Mercury raises $200M. The startup banking platform closed a $200 million Series D at a $5.2 billion valuation. It is betting on an AI-era financial operating system for founders. (Source: Tech Startups).

The week’s pattern is clear. AI is leaving the screen. That might be a Singapore AI testbed full of delivery robots, or an agent booking your apartment viewings. So if you are hiring across borders in APAC while your workforce shifts, Asanify’s global employer of record can help. It handles compliant hiring and payroll in new markets, so your team focuses on the work machines still cannot do.

FAQ: The Singapore AI Testbed and the Week in AI

What is the Singapore AI testbed at Punggol?

The Singapore AI testbed is a plan announced at ATxSummit 2026. The Punggol Digital District becomes a real-world site for robots. Companies like Certis, DHL, and Grab will deploy delivery, cleaning, and security machines in a public, mixed-use area. It is run by IMDA with JTC and the Singapore Institute of Technology.

How does physical AI affect HR and hiring?

As robots take over routine delivery, cleaning, and security tasks, frontline headcount shifts toward supervising and maintaining machines. Entry-level roles tend to get hit first. So HR teams need stronger upskilling and onboarding paths to keep junior talent employable.

What did Google announce at I/O 2026?

Google launched Gemini Omni Flash, a multimodal model that outputs editable video. It also made Gemini 3.5 Flash the default model behind AI Mode in Search. The update added agentic tasks that can act on a request over time. For example, AI Mode can keep working on an apartment search after a single prompt.

Not to be considered as tax, legal, financial or HR advice. Regulations change over time so please consult a lawyer, accountant  or Labour Law  expert for specific guidance.