In this written Q&A, the Bacalia Group founder explains why his investment outlook extends from AI and energy to space manufacturing, photonic quantum computing and brain–computer interfaces.
For Giuseppe Cicatelli, the question is not simply what comes after artificial intelligence. It is what becomes possible as intelligence grows more capable and widely available, and which constraints will matter next.
Cicatelli, Founder and Principal of Bacalia Group, discusses the commercial milestones he considers important, the sectors attracting his longer-term attention and why being early is not an investment thesis on its own.
Where do you see the next wave of investment taking shape?
I am less interested in asking what comes after AI than in understanding what becomes possible as intelligence gets cheaper, more capable and more widely deployed.
My starting point is the AI investment landscape: models, processors, computing infrastructure, applications and the software that makes a model useful. That includes the “harness” around AI: orchestration, tools, memory, evaluation and controls that allow it to perform tasks reliably.
Energy belongs in the same conversation, with nuclear among the options deserving close consideration. I also expect physical AI, including robotics and autonomous vehicles, to become an increasingly important destination for capital.
Further out, three areas particularly interest me: the space economy, quantum computing and brain–computer interfaces. I do not consider them undiscovered. I think their potential economic roles are still less widely understood than their most recognizable demonstrations.
What would make physical AI an attractive investment rather than an impressive demonstration?
Repeatable performance in a real operating environment. For software, I am interested in systems that can complete useful work reliably, not simply generate impressive answers. For a robot or autonomous vehicle, the test should be even stricter.
Maintenance, safety, supervision and cost all belong in the assessment. I would rather see a narrow industrial task with measurable economics than a spectacular demonstration without a clear customer.
I apply the same discipline to energy. Nuclear interests me, but a credible power-delivery plan is different from a technology narrative. Availability, construction, licensing, financing and the timing of demand all matter. A promising technology is not automatically an investable project.
Why do you see potential in space manufacturing?
I believe the space economy could become much larger as selected manufacturing processes move into orbit. The important word is “selected.”
I am not betting that ordinary factories will leave Earth. I am interested in high-value products for which the orbital environment could create an advantage large enough to justify transport, processing, recovery and regulatory costs.
Pharmaceutical processing is one area I would examine closely. My test would be whether a process performed in microgravity produces a meaningful advantage over what can be achieved on Earth, and whether that advantage survives the economics of the entire production chain.
A successful experiment would not, by itself, establish commercial-scale production or an approved medicine. The investment question is what can be produced repeatedly, to the required quality, at an economic cost.
Could that create opportunities beyond the products manufactured in space?
Yes, if those commercial conditions can be met. I would expect opportunities to develop around processing equipment, mission operations, reentry, recovery and quality assurance.
A manufacturing economy needs a reliable chain of services, not just a destination. The ability to produce something in orbit is only part of the proposition; the ability to return it dependably and integrate it into a customer’s operations matters too.
That is what attracts me to this area. Space could become part of a production process serving customers on Earth, rather than only a place where equipment is launched.
What interests you about quantum computing, particularly photonic approaches?
I am interested in approaches that connect scientific progress to a credible path toward manufacturing and operating much larger systems.
Photonic approaches use light to carry quantum information and can draw on semiconductor manufacturing techniques for their components. That creates an industrial path worth examining, but access to a manufacturing process does not, by itself, prove that a commercially useful, fault-tolerant system can be delivered.
It is important not to confuse photonic quantum computing with conventional computing that uses optical connections. The investment case I would examine concerns the ability to generate, manipulate and protect quantum information reliably at scale, not simply the use of light in a chip.
My thesis is not that quantum replaces classical computing or AI. It is that quantum could eventually provide a sufficiently valuable advantage for particular problems to support a durable commercial market.
How do you evaluate progress in a technology with a long development timeline?
I look at the path from component performance to a working system: reliability, error correction, manufacturing, integration and a customer’s reason to use it. An impressive chip is not, by itself, a complete investment thesis.
Patience must be distinguished from passivity. Long development timelines require milestones, financing discipline and a willingness to revise assumptions.
Believing that a technology will become important does not make every valuation or financing structure attractive. I want to understand what the next round of capital must accomplish and what evidence would cause me to reconsider the thesis.
Why are brain–computer interfaces another area of interest?
I see the potential for a much broader investment discussion, beginning with medical needs rather than speculative promises of human enhancement.
The applications I would prioritize are those that could help people with severe impairments communicate, interact with computers or control assistive devices. That gives the sector a concrete starting point: a clearly defined patient need, rather than an abstract promise to change the relationship between humans and machines.
I would not approach this as a single technology with one inevitable winner. Implantation method, signal quality, durability, clinical workflow and patient suitability all belong in the assessment. They are part of the product and, ultimately, its economics.
The question is not simply how much information an interface can capture. It is whether that information can be translated into a dependable benefit that patients and clinicians can use.
What would you need to see before the BCI opportunity becomes commercially compelling?
Meaningful patient benefit and a credible route through regulation, reimbursement and delivery. I would look beyond the implant itself to decoding software, clinical support, manufacturing and the practical requirements of long-term use.
Neural-data privacy, cybersecurity and informed consent also belong at the center of the assessment. These are not matters to address after adoption.
A compelling demonstration is not the same as a routine treatment, and a medical application should not be confused with a near-term mass-market consumer product. I would want the investment case to reflect those distinctions.
What could cause capital to move more decisively toward these sectors?
The common thread is not that they sound futuristic. It is that each could change a boundary around economic activity: where we manufacture, how we compute or how people interact with machines.
My working hypothesis is that attention could accelerate when technical milestones begin answering commercial questions. In space manufacturing, I would look for repeatable production, recovery and customer qualification. In quantum, progress toward useful, fault-tolerant systems. In BCI, durable clinical benefit and a practical path to serving patients.
Those milestones will not arrive on a synchronized timetable. I would resist treating the three sectors as a single investment cycle or assuming that visibility translates directly into returns.
The next wave may also reward less conspicuous companies. A supplier solving a difficult manufacturing, control or integration problem can be as interesting as the business attracting the headlines.
What is the biggest mistake investors could make when approaching these opportunities?
Treating “being early” as sufficient. My central conviction is that AI should expand our investment imagination, not consume all of it. But that does not make frontier investing forgiving.
Technical success may take longer than anticipated, require repeated financing or fail to produce attractive economics. An important sector can grow while individual investors lose money.
The question I keep returning to is simple: what must become true for this to be a useful business, and which part has already been demonstrated?
That is how I think about space manufacturing, quantum computing and brain–computer interfaces. Not as a forecast that every frontier will mature on schedule, but as a search for capabilities that could become indispensable before their economic importance is obvious.
