Cheap Cognition Is Just Phase One
Google DeepMind’s GNoME generated 2.2 million crystal structures in its search for new materials. The system identified 381,000 as predicted new, stable materials, and researchers found 736 that had been independently realized experimentally. 41 of those have been synthesized by an autonomous robotic laboratory.
That’s 800 years of science, compressed.
We are seeing the same pattern in cognitive work.
The cost of producing cognitive work is falling rapidly. Software, analysis, research, design and customer service can all be produced with much less human effort than before.
I see the economic effects separating into four categories.
1\ MORE OF WHAT WE ALREADY DO
Microsoft CEO Satya Nadella has described AI as changing the amount of work individual employees can handle. Google CEO Sundar Pichai has reported similar gains in software development.
Amazon CEO Andy Jassy gave a particularly concrete example. Amazon used AI coding tools to migrate 30,000 internal applications to a newer version of Java. The company estimated that the work saved 4,500 developer-years and $260 million in annualized costs.
IKEA has taken a similar approach in customer service. Its Billie AI assistant handles routine customer enquiries. The company has retrained roughly 8,500 customer-service employees for interior-design advisory roles creating a new billion-dollar business.
The underlying activities remain familiar.
Customers need answers. Software needs maintenance. Employees need to work through documents and data.
AI changes the amount of human effort required.
A developer can spend less time on repetitive migration work. A customer-service organization can handle more enquiries. An analyst can process more documents.
This is the productivity effect.
2\ THINGS WE KNEW WERE POSSIBLE, BUT DIDN’T DO
Every organization has a large number of activities sitting below its economic threshold.
A ten-person company might benefit from software designed around its own operations. A retailer might like to analyze every customer individually. A research team might have hundreds of interesting questions that would each take several weeks to investigate.
The work can be done.
The question has always been whether it deserves the people, time and money.
AI changes that calculation.
Microsoft’s 2026 Work Trend Index describes companies using AI to give employees greater capacity and redesigning roles around people working with AI systems.
Consider a small manufacturer that has always wanted a production-planning system tailored to its factory. A conventional software project might cost hundreds of thousands of dollars and take months. With AI doing much of the specification, coding, testing and documentation, the same company can approach the project very differently.
The same calculation applies to research, customer analysis, translation, legal work and thousands of specialized services.
There are many things businesses would like to do if the cognitive labor were cheap enough.
Some of them are now becoming economical.
3\ THINGS THAT WERE ALREADY THERE, BUT WE HADN’T FOUND
The third category is discovery.
There are enormous quantities of information that humans have already collected and enormous spaces of possibilities that humans have barely searched.
Astronomy provides a wonderful example.
In January 2026, the European Space Agency announced that researchers had used AI to examine almost 100 million image cutouts from the Hubble archive. The system completed the search in about two and a half days and identified nearly 1,400 anomalous objects. More than 800 had never been documented before.
The objects were already in the images. The archive had been sitting there for years. The new capability was the ability to examine it at a scale that people could not.
GNoME works on a similar principle. There are enormous numbers of possible crystal structures. The system searched through a space that would have been impractical for human researchers to explore comprehensively.
Mathematics has the same problem. So do drug discovery, protein design, engineering and many areas of physics. And we are seeing incredible solutions emerging across these spaces.
The world contains more possibilities than we have historically had the capacity to examine. I wrote about this recently on incredible breakthroughs in math from OpenAI, AI designing new bacteriophages (viruses capable of overcoming drug-resistant bacteria), and more.
Cheap cognition changes the size of the search we can afford.
That means discoveries can come from places that have been sitting in front of us for years.
4\ THINGS WE CREATE FOR THE FIRST TIME
A human and an AI system can work through a design space together and arrive somewhere neither had previously considered.
That could be a material, a product, a piece of software, a scientific method, or an entirely new business.
Film is beginning to provide early examples.
In 2026, the AI-focused studio Promise began producing Touch Grass, using AI-generated environments and other synthetic production elements. The production approach changes the economics of creating locations and visual worlds that would traditionally require sets, physical locations or large visual-effects teams.
The studio has attracted investment from companies including Google and Disney.
Here are some incredible examples from Higgsfield’s Original Series:
In the coming months watch for evidence of this across sectors.
An engineer will generate a design that nobody had considered. A scientist will develop a hypothesis from an unusual combination of evidence. A software team will build a product for a market that previously looked too small. A company will discover a way to operate that would have required a much larger organization.
The combination of human judgment, machine search and rapid iteration gives us access to a much larger creative process.
That is where the possibility space itself can start to change.We have spent the first phase of AI making cognition cheap.
Now we can do more of what we already do. We can afford activities that were previously too expensive. We can search possibilities we could never examine at scale. AND we can create things that would never have been attempted before.
In my books, cheap cognition was phase one. What comes next will be FAR more interesting.
Until next time,
Ram
—
Ram Srinivasan
MIT Alum | Author, The Conscious Machine | Global Future of Work and AI Adoption Leader published in Business Insider, Fortune, Harvard Business Review, MIT Executive Viewpoints and more.
—
A Message From Ram:
My mission is to illuminate the path toward humanity's exponential future. If you're a leader, innovator, or changemaker passionate about leveraging breakthrough technologies to create unprecedented positive impact, you're in the right place. If you know others who share this vision, please share these insights. Together, we can accelerate the trajectory of human progress.
Disclaimer:
Ram Srinivasan currently serves as an Innovation Strategist and Transformation Leader, authoring groundbreaking works including "The Conscious Machine" and the upcoming "The Substrate Shift."
All views expressed on "Substrate" and across all digital channels and social media platforms are strictly personal opinions and do not represent the official positions of any organizations or entities I am affiliated with, past or present. The content shared is for informational and inspirational purposes only. These perspectives are my own and should not be construed as professional, legal, financial, technical, or strategic advice. Any decisions made based on this information are solely the responsibility of the reader.
While I strive to ensure accuracy and timeliness in all communications, the rapid pace of technological change means that some information may become outdated. I encourage readers to conduct their own due diligence and seek appropriate professional advice for their specific circumstances.
