Humans in the Loop

NVIDIA CEO took some time in the last week to posit his predictions for the long term future of AI. From AI creating more millionaires in the next 5 years than the internet has done since it launched, to AI generating more jobs than it kills, he was fairly bullish on the matter.

Huang is a very important voice in the field as the leader of chip company NVIDIA. He transcends both business and politics. He’s one of the few – at least at the time if writing – that is able to help President Trump see value in the US-China relations.

He is also a visionary thinker. I found his speculation on jobs particularly interesting. Before that, below is a nice 3×3 grid of his headline thoughts followed up by a bit more of my take on Huang’s speculative thoughts.

James Stewart

AI as core infrastructure
Every large company will treat AI as essential as electricity or the internet, building “AI factories” (advanced data centres) to generate valuable outputs.

Massive wealth creation
Small, focused AI research teams will drive unprecedented economic value, spawning a wealth boom that outpaces even the internet’s impact. You only have to look at start-ups like Lovable for that to seem true today.

Twin operations for every business
Organisations will run both a physical operation and an AI “twin” that continuously optimises processes in real time. This means dual running factories, designing, prototyping and producing at exceptional speed and innovation.

Trillion‑dollar infrastructure boom
Annual spend on AI data centres is set to exceed US$1 trillion by 2028, marking the next major wave of industrial investment.

Rise of agentic AI
Autonomous AI agents, “information robots” that perceive, plan and act. They will transform workflows and create a multi‑trillion‑dollar industry. The recent announcement by OpenAI is a near reality of this already happening.

Autonomous vehicles at scale
Self‑driving will become a dominant form of transport, with a large share of global miles driven autonomously. My teenage kids won’t need to learn to drive, maybe?

Job disruption and re‑creation
AI will both eliminate routine tasks and boost overall productivity; while some roles disappear, new jobs will emerge as innovation accelerates. Huang even said his company, despite being 100% AI adopters, have more work than resources. Their backlog is significant.

Continued support for open‑source AI
Broad, community‑driven models will democratise AI access. Especially in emerging economies, fueling global innovation. There are some unicorns going to rise out of non-traditional spaces.

Physical AI for real‑world tasks
Investments in robotics and autonomous systems (e.g. physics engines like Nvidia’s Newton) will bring AI into manufacturing, logistics and beyond. Once the motor skills of a robot get close to mimic human levels this will revolutionise much of the mundane work done in a factories that is unsafe and repetitive.