Google published an interview this week with three DeepMind researchers behind Gemini Omni, its model for generating and editing video from any input. The interview is genuinely fun: swapping hairstyles, turning people into sloths, materialising a cat on a table. "There is no ceiling to this," one researcher said. Fair enough. It is impressive technology.
It is also a useful moment to separate two questions Australian business owners tend to conflate: is this technology impressive, and does my business need it right now. Those are not the same question, and answering the first as if it settles the second is how companies end up spending money on capability they never use.
Novelty demos are not a roadmap
Every few months a new generative model ships with a viral demo reel, and every few months a wave of businesses feel behind for not having a plan for it. Almost none of them actually need one. Consumer-facing video generation is a real and growing category, marketing content, social assets, product previews, but it is a specific tool for a specific job, not a general business change. The reel is designed to impress, not to map onto your P&L.
The businesses we see get real value from AI in 2026 are not the ones chasing every new generative-media release. They are the ones with a short, unglamorous list:
A clear, boring problem that costs real time or money: invoice processing, lead triage, customer follow-up.
A model matched to that problem's stakes, not the newest or flashiest one available.
A way to measure whether it is actually working, beyond it looked good in a demo.
The cost of chasing the reel
Chasing hype is not free even when the tool is cheap. A small Australian firm that decides it needs an AI video strategy typically spends on tooling subscriptions, a contractor to learn the tool, and management time debating brief and output. It is easy to sink $5,000 and a month of attention into a capability that produces a handful of clips nobody was waiting for, while the invoicing process that quietly costs a day a week goes untouched. The opportunity cost is the real bill: every hour spent on the exciting thing is an hour not spent on the expensive, boring thing that would have paid for itself.
Where Claude fits into this
Anthropic has not shipped a splashy video-generation demo, and that is a deliberate choice, not a gap. Claude's roadmap has stayed anchored on the categories where Australian businesses actually spend money: coding and agent reliability, document and data work, and now proactive team collaboration through Claude Tag. None of that generates a viral clip. All of it shows up in hours saved and errors avoided, which is the metric that matters when you are the one signing off the budget.
What not to conclude
This is not a claim that AI video is a fad or that Gemini Omni is not worth having. If your business genuinely sells on visual content, an agency, an e-commerce brand, a media team, these tools are worth evaluating on their merits, and dismissing them would be its own mistake. The point is narrower: let the job decide the tool. If you are feeling pressure to have an AI video strategy because a demo looked exciting, that pressure is not a business case. Solve one specific, expensive, boring problem well before you buy the thing that photographs nicely.
How to tell hype from a real need
There is a quick test that cuts through most of it. Ask who is asking for this capability, and what breaks if you never build it. If the honest answer is that a competitor posted a slick clip and it made you nervous, that is not a need, it is a feeling, and feelings are expensive to build for. If the answer is that a paying customer keeps asking for video assets you currently outsource at real cost, now you have a case worth costing. The reel did not create that demand; the customer did, and that is the difference between a strategy and a reaction.
The same test works across every hype cycle, not just video. When the next viral model lands, and it will within weeks, run it through the same two questions before you touch the budget. Most of the time the answer is to keep your attention on the expensive, boring work already in front of you, and let the demo stay a demo until a real customer turns it into a requirement.
If you want help working out which category your business actually needs, before you spend a dollar on any of it, book a session and we will start from the problem, not the demo.
