The closest modern matches are not isolated inventions. They are systems whose key components already exist but have only recently become good and cheap enough to combine into mass markets.
The strongest candidates are AI, robotics, engineered biology, and electrified energy systems. Each resembles the early automobile or computer industry in a different way.
1. Foundation-model AI
This is the clearest example—and may already be past the “emerging” stage.
What converged
- large neural-network architectures
- specialized chips and cloud data centers
- enormous digital datasets
- cheap storage and networking
- better training methods
- natural-language interfaces
- software distribution to billions of devices
Individual elements existed for decades. The breakthrough came when scale made the whole system useful.
Its early market
AI first flourished in places willing to tolerate cost and imperfection:
- advertising and recommendations
- translation and transcription
- programming assistance
- customer support
- media generation
- research and document analysis
These markets finance further improvement, much as wealthy automobile buyers funded the industry before mass ownership.
Why it may be transformative
Computers reduced the cost of executing formal instructions. AI reduces the cost of producing certain kinds of judgment, language, code, design, and analysis.
That distinction matters. Traditional software automates a process after a programmer specifies the rules. AI can automate tasks for which the rules are difficult to state explicitly.
The missing infrastructure
AI still needs its equivalents of roads, traffic laws, and filling stations:
- reliable evaluation
- identity and authorization systems
- better organizational data
- liability rules
- workflow redesign
- inference infrastructure
- tools that make models dependable, auditable, and secure
The largest businesses may not be model creators. They may be companies that reorganize entire workflows around abundant machine intelligence.
2. General-purpose robotics
Robotics looks like the automobile industry before reliable mass production: impressive demonstrations, valuable specialized machines, but limited general adoption.
What is converging
- computer vision
- language and multimodal models
- cheaper sensors
- better batteries
- improved motors and actuators
- simulation and synthetic training data
- falling compute costs
- more flexible manufacturing
AI gives robots something they historically lacked: a way to interpret ambiguous instructions and cope with environments that were not completely programmed in advance.
Likely initial markets
General-purpose robots will not begin by doing everything in a household. They will start where environments are constrained and labor is expensive:
- warehouses
- factories
- hospitals
- agriculture
- mining
- construction
- defense
- commercial cleaning
These are the robotics equivalent of early commercial computing. A machine can be expensive and imperfect if it eliminates enough dangerous, repetitive, or scarce labor.
Why it may be transformative
AI without robotics mostly changes information work. Robotics could bring the same cost decline to physical work.
The important threshold is not “a robot that looks human.” It is:
Can a machine perform enough valuable tasks per day to justify its purchase, maintenance, supervision, and downtime?
Once that answer becomes yes in one narrow setting, production volume can lower hardware costs and expand the next market. That is how an industry compounds.
3. Engineered biology
Synthetic biology may be the closest conceptual parallel to computing. DNA is not literally software, but biology increasingly behaves like an engineerable information system.
What is converging
- cheap DNA sequencing
- gene editing
- DNA synthesis
- automated laboratories
- protein-structure prediction
- AI-assisted molecule design
- large biological datasets
- improved cell and gene delivery
Sequencing gives us the ability to read biology. Gene editing and synthesis let us modify or write it. Automation and AI accelerate the design-test cycle.
Early markets
The first successful markets tolerate high costs because the value is unusually high:
- rare-disease treatments
- cancer therapies
- vaccines
- high-value pharmaceuticals
- research tools
- specialized enzymes
This resembles early computers, which were economically rational for military, scientific, and large corporate users long before they were affordable for households.
Potential mass markets
If costs and reliability improve, engineered biology could spread into:
- common-disease prevention
- personalized medicine
- cultivated or fermented foods
- industrial materials
- chemicals made without petroleum
- climate-resilient crops
- biological manufacturing
Main constraint
Biology is harder than software because cells are noisy, evolved systems rather than designed machines. Experiments take time, manufacturing is difficult, and failures can harm patients or ecosystems. Progress may therefore be slower and less predictable than software enthusiasts expect.
4. Batteries and the electrification of machines
Batteries can look merely incremental: the same old devices with a different energy source. But electric motors, power electronics, batteries, and software together could reorganize transportation and energy.
What is converging
- falling battery costs
- better energy density and lifetime
- power electronics
- renewable generation
- charging infrastructure
- grid software
- electric motors
- large-scale manufacturing
Markets already opening
- passenger vehicles
- delivery fleets
- buses
- scooters and bicycles
- lawn and construction equipment
- home energy storage
- grid balancing
The deeper opportunity is not just replacing gasoline cars. Electrification makes machines simpler, more controllable, and easier to integrate with software.
Why it resembles automobiles
A reinforcing system is forming:
- More electric products increase battery production.
- Scale reduces battery costs.
- Lower costs open new categories.
- More categories justify charging and grid investment.
- Better infrastructure increases adoption.
This is the same kind of feedback loop that connected cars, petroleum, roads, suburbs, and mass production.
5. Autonomous vehicles and machines
Autonomy is best understood as a layer built from AI, robotics, electrification, maps, sensors, and communications—not as a standalone invention.
Its first large markets are likely to be situations in which routes or environments are limited:
- mines and ports
- warehouse vehicles
- highway trucking
- fixed-area taxis
- agricultural equipment
- drones
- defense systems
The common mistake is to ask when one system will drive everywhere under all conditions. Early automobiles could not travel everywhere either. Industries scale from economically valuable subsets, not from instant universality.
Autonomy could change the economics of transport because vehicles are currently underused and labor is a large operating cost. But regulation, safety, insurance, and edge cases make the timeline uncertain.
6. Cheap launch and commercial space infrastructure
Reusable rockets are reducing the cost and increasing the frequency of reaching orbit. That alone does not guarantee an automobile-sized industry, but it changes what businesses are possible.
Enabling stack
- reusable launch vehicles
- improved engines and manufacturing
- small satellites
- standardized satellite components
- cheaper sensors
- ground-station networks
- high-volume launch operations
Early markets
- communications
- Earth observation
- navigation
- defense
- scientific missions
The most credible near-term transformation is not asteroid mining or Mars settlement. It is treating orbit as infrastructure: communications, sensing, positioning, and perhaps manufacturing. Its impact could be enormous while remaining largely invisible to consumers.
7. Augmented-reality and spatial-computing devices
This category has long looked imminent and repeatedly disappointed. Yet it fits the convergence pattern:
- compact displays
- computer vision
- low-power chips
- spatial mapping
- batteries
- wireless connectivity
- voice and AI interfaces
The likely mistake is assuming the winning product must begin as an all-day consumer headset. Better beachheads may include:
- industrial maintenance
- medical procedures
- logistics
- training
- defense
- remote technical assistance
AI could make the interface more valuable because users can ask a system about what they are seeing rather than navigate conventional menus.
Still, this is a weaker candidate than AI or robotics. Comfort, social acceptance, battery life, and unclear everyday value remain major barriers.
A useful way to rank them
| Candidate |
New capability created |
Cost decline underway |
Initial market exists |
Infrastructure forming |
Transformational potential |
| Foundation-model AI |
Very high |
High |
Yes |
Yes |
Very high |
| General robotics |
Very high |
Moderate |
Yes, narrowly |
Early |
Very high |
| Engineered biology |
Very high |
High in tools |
Yes |
Yes |
Very high |
| Electrification |
Moderate |
High |
Yes |
Yes |
High |
| Autonomous systems |
High |
Moderate |
Yes, narrowly |
Early |
High |
| Commercial space |
Moderate |
High |
Yes |
Early |
High but uncertain |
| Spatial computing |
Moderate |
Moderate |
Narrow |
Early |
Uncertain |
These industries will also combine. AI may become the control layer for robotics, biological design, autonomous transport, and energy management. Batteries may power robots and autonomous vehicles. Space systems may provide communications and observation to all of them.
That interaction is often where the largest effects appear.
How to recognize the pattern early
Look for six conditions.
1. A formerly scarce capability is becoming abundant
Examples include intelligence, physical automation, biological design, energy storage, or access to orbit.
2. Several cost curves are improving together
One improving component creates a product. Five improving components can create an industry.
3. An expensive but forgiving beachhead exists
The first customers usually care more about performance than price. Hospitals, militaries, factories, researchers, and large enterprises often play this role.
4. Use creates more improvement
Powerful industries have feedback loops:
- more AI use generates better data and tools
- more robots create manufacturing scale
- more batteries reduce battery costs
- more sequencing improves biological knowledge
5. Complementary infrastructure is appearing
The less glamorous complements—insurance, standards, maintenance, financing, training, regulation, distribution—often indicate that a technology is becoming an industry.
6. Entrepreneurs are moving from replacement to redesign
The earliest products imitate what came before: cars resembled carriages, and early web pages resembled printed documents. The deeper transformation begins when people redesign institutions around the new capability.
The strongest current answer is therefore AI combined with robotics. AI alone could become as consequential as computing; robotics could extend it into the physical economy. Engineered biology is the next strongest candidate, though its development cycle will probably be slower. Electrification is less conceptually novel, but it may produce an equally large industrial transformation.
The key is not to search only for a spectacular object. Search for a capability whose cost is collapsing, then ask what new system becomes possible when businesses, infrastructure, and human behavior reorganize around it.