When I look at the technology industry today, it is difficult to describe it as a mature market that has simply settled into place.
The opposite seems to be happening.
Since 2020, the U.S. technology sector has gone through a remarkable period of new company formation, private investment, AI development and infrastructure spending. The biggest change has been the rise of artificial intelligence, but the impact goes well beyond AI models themselves.
New companies are being created. Venture capital is moving into areas such as AI, robotics, semiconductors and data centres. Hyperscalers are spending enormous amounts on infrastructure. And established technology markets are seeing new competitors emerge surprisingly quickly.
For me, the interesting question isn't simply how much money is being invested.
It is what that investment is producing.
New Businesses Are Still Being Created at a High Rate
One of the clearest ways to measure entrepreneurial activity is to look at applications to start new businesses.
U.S. Census Bureau data cited by the Computer & Communications Industry Association (CCIA) shows that new business applications have remained well above their pre-2020 levels. The source reports nearly 524,000 applications in May 2026, while applications from businesses considered likely to become employers were also significantly higher than before 2020.
That matters because the technology story isn't being driven exclusively by a handful of established companies.
Some of the most valuable new technology businesses were barely known a few years ago.
OpenAI was founded in 2015. Anthropic was founded in 2021. xAI was founded in 2023.
Their rapid growth illustrates how quickly a new company can move when a new technological platform creates a large market opportunity.
AI Has Become the Centre of the Investment Boom
AI is clearly at the centre of this change.
According to the CCIA analysis, U.S. venture capital investment reached $412.7 billion during the first half of 2026, exceeding the total for the entire previous year. The analysis says approximately 86% of that first-half investment went to AI companies.
Those numbers are enormous, but they also need some context.
Not every dollar going into an AI company represents the same thing.
Some funding goes toward model development. Some supports semiconductor companies. Some goes into robotics and autonomous systems. Other investment is directed toward data centres, networking infrastructure and applications built on top of existing models.
In other words, the AI economy is becoming a much larger ecosystem.
The Infrastructure Behind AI Is Expanding Too
This is probably the part of the story that is easiest to overlook.
AI requires physical infrastructure.
Models need computing power. Computing requires chips, servers, networking equipment, electricity and data-centre capacity.
The four largest U.S. hyperscalers reportedly spent around $410 billion on capital expenditure in 2025, with company guidance pointing toward roughly $700 billion in 2026. The spending is heavily connected to AI data centres, chips and networking infrastructure.
That is a major shift from the way we normally think about software companies.
AI may be delivered through software, but supporting that software requires a very physical economy.
Data centres have to be constructed.
Power has to be generated and transmitted.
Servers have to be manufactured and installed.
Networks have to handle increasing volumes of data.
The AI boom is therefore creating demand across several traditional industries at the same time.
Data Centres Are Becoming an Economic Story
The construction numbers help make this more tangible.
According to Census Bureau data cited by CCIA, private data-centre construction reached an annualised rate of $50.7 billion in April 2026, up 28% from a year earlier.
The underlying trend is even more striking. Real annual data-centre construction spending more than tripled between 2019 and 2025, according to the analysis.
I think this is important because data centres are no longer just an infrastructure issue for technology companies.
They are becoming part of the broader investment and economic picture.
The AI industry needs them, cloud companies need them, enterprises increasingly depend on them, and governments have to think about the energy and infrastructure requirements that come with them.
New Entrants Are Changing the Competitive Landscape
Another thing I find interesting is how quickly the competitive environment can change.
AI has created space for companies that didn't exist at the beginning of the current generative-AI cycle.
OpenAI's ChatGPT became a mainstream consumer product after launching in late 2022. Anthropic entered the market later, while other companies have emerged around AI models, chips, robotics and specialised applications.
The CCIA analysis points to rapid changes in enterprise AI spending shares as one example of how quickly competition is developing. It cites Menlo Ventures estimates showing OpenAI's share of enterprise LLM spending falling from 50% in 2023 to 27% in 2025, while Anthropic and Google gained share.
Those figures are estimates rather than a complete measurement of the entire
0 Comments