The AI Excitement in Silicon Valley Is Unlike Anything Since the Cloud Era

There is an energy in the technology industry right now that I have not witnessed since the emergence of cloud computing. Silicon Valley has recognised that the entire digital economy is changing at breakneck speed, and the ecosystem is building supporting infrastructure accordingly.

Most of the new startups I encountered at HumanX operate in this space. Whether it is evaluating model outputs (e.g. Braintrust), orchestrating agents (Glean), or penetration testing agents (ZioSec), there are dozens of companies helping organisations mature their AI footprint from a technology perspective. The other major area of investment is agents that actually do things. As the AI stack matures and trust in model outputs increases, agents will move up the stack into higher-impact, higher-value use cases.

My read of where we are today: agents in production, and genuinely productive, are still relatively scarce (as also pointed out in an MIT study stating that 95% of AI experiments don’t make it to production), with the exception of sales and software development. I expect broader production workloads to emerge quickly.

Why US Companies Are Pulling Ahead of Australia and Europe on AI Adoption

Where many Australian businesses still appear to be treating AI as an advanced search engine, a subset of existing technology capabilities, or even a meaningless bubble, American businesses are actively investing in Head of AI roles and building frameworks to scale their productivity. My personal view is that the Head of AI will be the one of the most important roles in an organisation for at least the next five years.

American businesses are also simply moving. Some companies that are all-in on cloud infrastructure expect their AI spend to surpass their cloud spend within the next six months. Based on first-hand conversations at HumanX, my estimate is that both Australia and Europe are roughly six to twelve months behind Silicon Valley. OPTIMAZE was the only Australian company at HumanX.

Will AI Replace Enterprise Software? What the Future of the Software Stack Actually Looks Like

Large enterprise software packages that hold proprietary data, complex workflows, and deep industry expertise will continue to be valuable for the immediate future. While there are use cases at the edges of these products that are replaced by self-built or lightweight AI-powered alternatives, full replacement of core systems will take time. Perhaps that is an argument against the so-called SaaSpocalypse, but they can't sit still though.

What will change, however, is how we engage with enterprise products. More and more software products will be consumed directly through the LLM provider of choice. This is why OPTIMAZE is investing in being present natively in these tools. It is a prime example of the existing stack, think Microsoft Windows and Apple OS plus the browser/app, being reinvented from the ground up. It would not surprise me if the large LLM companies eventually release a fully AI-native operating system that becomes your digital twin throughout your life.

How Australia Can Close the AI Gap Before It Becomes Permanent

We are in a race. Every part of the digital and physical world is being reimagined with AI. While not all of it will reach production immediately, the pace at which models are improving and becoming more fine-tuned means some of today’s experiments may become profitable with just a single model update, or switch. The window to catch up is open, but it will not stay open indefinitely.

How to Know If You’re Investing in the Right AI Initiatives (And Avoid Wasting 95% of Your Budget ) 

So how do you know where to invest, and whether you are backing the right initiatives? You measure it.

The race will not be won by spending the most on AI. Knowing exactly where to double down is critical to avoid wasting the majority of your capital on AI that is not yielding results. An MIT study found that companies without proper measurement frameworks waste up to 95% of their AI spend on initiatives that produce no meaningful return. This is precisely where OPTIMAZE helps every company make better AI investment decisions, protecting margin and profitability in the process.

Now Is the Best Time in History to Build Something New

If you have ever wanted to launch an initiative, the time is now. The barrier to entry for building digital products has effectively disappeared. What remains is execution.

OPTIMAZE partners with companies on that journey, helping you double down on the features and ideas that generate real ROI, and cut the ones that do not.

OPTIMAZE was a sponsor at the HumanX AI conference in San Francisco, April 2026.