San Francisco – April 15, 2026

By Emilia Pasquier, CEO of Swissnex in San Francisco

Last month, I headed east across the Atlantic to meet with Switzerland’s innovation ecosystem at Innoday 2026, the national conference on the public promotion of Innovation in Switzerland. I presented some trends that my colleague Maulde Cuérel and I have been observing in Silicon Valley to an audience of Innovation specialists.

Silicon Valley, and its neighbor to the north, San Francisco is all about new trends, and about discovering what might be the next innovation to disrupt specific parts of society. What happens there is often a reliable peek into what the future could look like, not only in the U.S., but also in Switzerland. Some of these innovations do end up changing our societies entirely, the most famous examples being the platform economy and social media. Other ideas have their moment of fame but have not (yet) delivered on their promises to revolutionize the world. Think smart glasses or the metaverse.

I’ve lived in San Francisco for three and a half years — almost from the moment ChatGPT went public — and I have seen the city changing from its post-covid rather gloomy phase to a new – maybe not goldrush – but AI-rush. In the midst of all the noise around AI, Maulde and I have tried to extract the main three trends that we believe are not merely temporary obsessions, but longlasting trends that are profoundly changing our societies.

1. The Newcomers — Autonomous systems and robotics

Most of you know that autonomous cars have already been around for a while in San Francisco. There have been accidents, such as the widely reported 2023 incident involving Cruise, where an autonomous car dragged an injured woman more than 20 feet. The vehicle had initially stopped, but its sensor system failed to detect her beneath it. Cruise’s handling of the incident led to its indefinite suspension in California.

This accident could have been a complete defeat for autonomous vehicles in San Francisco, but other companies — such as Waymo, owned by Alphabet, and Zoox, owned by Amazon — continued to deploy their cars in the city. And I can share my own experience, which mirrors what I hear from other users: we generally feel safer in a self-driving car than in a human-driven one. The autonomous car doesn’t answer calls while driving and doesn’t look for its sunglasses during the ride. For instance, I ride my bike to the office every day, and when I encounter a human-driven car, I slow down, make sure to make eye contact with the driver, and keep my hands on the brakes in case the driver makes an unexpected move. When I see a robotaxi, I generally ride at full speed, knowing it will ‘see’ me. My own experience of adapting to these new road users leads me to believe it is only a matter of time before autonomous vehicles become a standard part of everyday traffic.

The numbers point in that direction. Waymo recently closed a massive $16 billion funding round boosting its valuation to $126 billion. It also continues to scale its operations, expanding beyond the U.S. testing to cities like London and Tokyo.

The revolution might not only happen in cities. Autonomous trucking is also on the rise. It offers a simpler operating environment: highways, more predictable conditions, less complex navigation, and clearer economics. Companies like Aurora are already testing long-haul routes that can operate far longer than human drivers are legally allowed to. This is where autonomy may scale faster than expected.

2. The Shift — AI for military and defense

For a long time, Google’s motto “don’t be evil” reflected a broader reluctance across tech to engage with politics — and even more so with defense. That line is moving. Major tech companies are now increasingly open to their technologies being used in military contexts. Anthropic was holding an USD200 millions contract with the Pentagon, but the relationship collapsed in February 2026 when the Pentagon demanded unrestricted use of the technology — including for mass surveillance and autonomous weapons — and Anthropic refused. This opened an opportunity for their competitors, and OpenAI stepped in and signed the deal.

It is true that the boundary between civilian and defense technology has always been a grey one. But today, the difference might be that we see more events labelled “’defense tech”. Innovations such as advanced computing, cybersecurity, biotechnology, and autonomous systems are the same ones driving mainstream commercial innovation. The same software that powers a logistics startup can guide a drone; the same AI that recommends your next purchase can analyze a battlefield. Investors are taking notice: in January 2026, Andreessen Horowitz closed a nearly $1.2 billion fund dedicated to defense technologies, a strong signal that Silicon Valley is no longer shy about where this money is going.

The U.S. Department of Defense has accelerated its work on AI, publishing strategies and funding projects with the clear objective of maintaining technological leadership. Their direction is clear: AI is no longer just a commercial race — it is a strategic one.

3. The Bottleneck — Energy

The CEO of Nvidia, Jensen Huang, gave this impressive stat during the keynote he held at the major company’s annual event Nvidia GTC: in the last two years, the demand for compute power has been multiplied by one million. Whether or not his number is exact or not, the underlying reality is hard to ignore: the demand for compute, and therefore energy, is exploding.

While training large models was the primary driver of AI energy consumption in recent years, inference (the computational work of running AI models) has now become the dominant source of demand. As AI is adopted by more people and integrated into more areas of life, this pressure on energy systems will only compound. What is interesting in the public discourse is that there seems to be a growing recognition, globally, that energy could become a constraint and that AI companies might not be able to scale at the speed they envisioned.

Approaches to this challenge, however, differ around the world. Where Europeans often see regulation as part of the solution to the exploding energy demand, Silicon Valley believes in progress over regulations. I can hear its relentless optimism at the events I am attending: models will become smaller, energy cheaper, data centers more efficient.

This faith in progress is visible in the current Silicon Valley bets to solve the energy constraints, such as the U.S.-based Helion Energy, a fusion energy startup, backed by Sam Altman, or in more unconventional ideas explored by Elon Musk around rethinking infrastructure at scale and sending data centers into space to cool them.

3.5 AI with hands — The Agent

Three years for three trends, and half a year for half a trend: the agent. I believe agentic AI will profoundly reshape how we live, work, and organize our lives — but since it is only half a trend for now, I will save it for the next issue. Stay tuned!

This is part of a newsletter published on LinkedIn. Subscribe here to get alerted about new articles.