The startup landscape is shifting towards hard tech, with a growing number of companies focusing on physical products enabled by AI advancements that accelerate research and reduce engineering barriers. Simultaneously, software startups are evolving with AI-powered full-stack solutions and data-driven approaches, while experienced solo founders leverage AI tools to innovate and scale rapidly across both physical and digital domains.
The recent surge in hard tech startups marks a significant shift in the startup landscape, with companies increasingly focusing on physical products that interact with atoms rather than just software. At Y Combinator (YC), the proportion of hard tech companies has grown from 8% to 20%, with notable growth in sectors like robotics, industrial manufacturing, defense, semiconductor technology, and power infrastructure. This resurgence is fueled by advancements in AI and machine learning, which accelerate scientific research and product development, enabling startups to achieve breakthroughs faster and more efficiently than before.
A key driver behind this trend is the influx of highly technical founders, many holding PhDs, who bring deep expertise to complex fields such as silicon photonics and robotics. AI tools, particularly advanced coding agents, have reduced the need for large engineering teams by automating significant portions of software development, thereby lowering the barriers to entry for hard tech ventures. Additionally, macro trends such as the success of space-related companies like SpaceX, increased defense tech innovation, and the growing demand for AI compute power have created fertile ground for startups addressing these physical and technical challenges.
Alongside hard tech, software startups are evolving, particularly those building full-stack, end-to-end solutions that automate entire workflows rather than offering point solutions. These AI-powered products are demonstrating rapid revenue growth, with some startups reaching seven-figure revenues within just three months of YC batches—an acceleration from the typical 18-month timeline. Companies like Juicebox exemplify this shift by integrating AI agents that not only assist but actively perform tasks such as recruiting outreach, thereby enhancing productivity and customer value.
Another emerging and lucrative category is companies providing data and reinforcement learning (RL) environments to AI labs. These startups supply critical training data and simulation environments essential for advancing AI models, with some generating hundreds of millions in revenue within a few years. This data-driven approach is especially important for robotics, where customized models trained on specific environments enable real-time, precise physical interactions, highlighting the growing intersection of AI and physical intelligence.
Finally, there is a notable resurgence of experienced founders in the startup ecosystem, with a significant increase in solo founders who possess the technical and business acumen to build and scale companies independently. The availability of powerful AI coding agents has democratized the ability to develop complex products, reducing reliance on large founding teams. This shift encourages more seasoned entrepreneurs to leverage their expertise and manage AI-driven development teams effectively. The overall message is clear: the convergence of AI advancements and renewed interest in hard tech is creating unprecedented opportunities for builders to innovate and scale rapidly in both software and physical domains.