I have long enough been in this business to realize when it is nothing but empty talk. New coin, new tale, old idea in other words. Therefore I did not even mind about the angle of the robot when I first saw Fabric. We've already seen that movie.
There was another thing that drew my attention. Fabric is attempting to make rules to machines
Real digital guardrails
Listen, here’s the thing. Robots nowadays do not operate in a common mechanism. They remain within the company walls, doing anything they are ordered to do and there is no definite means of establishing accountability. When something goes wrong, then it is difficult to trace. It becomes emails, support tickets and blame games. The chaotic truth is that, there is still no actual legal system of machines.
Fabric is trying to write one.
And it is not doing it in an abstract manner. The system itself consists of the rules. Smart contracts define the duties. Conditions are clear. Outcomes are predictable. A machine is either able to fulfill the requirement or not. No gray area, or, at any rate, less of it.
That change is more than an unsuspecting human being opens his eyes to see. Since you cannot depend on human beings to make all the decisions since once machines begin to act more autonomously. You must have a framework that does not require monitoring. Instead of management layer, something like a rule engine.
I believe that this is the actual bottleneck
Fabric relates this rule system to incentives. Not just in theory. In a most simple, most tough way. You stake value to join. When you do the right thing, you make money. If you don’t, you lose. No appeals. No ‘we’ll fix it later.’ The system does not need to be administered manually.
That is why it is not that much of a tech product but infrastructure. Or perhaps as an in-built machine contract system. Not people to people contracts. Contracts between machines.
And then there’s governance. This aspect is normally overlooked or buried under buzzwords yet it counts. In case machines are to run at scale, someone or something must make a decision on how the rules vary. Fabric makes that an open system, rather than keeping it a company dashboard secret. It is disheveled, yet likely it was necessary.
Still, none of this is clean.
It is all great to build a rule system of machines until you encounter real-world complexity. Sensors fail. Data gets noisy. Edge cases pile up fast. It is already difficult to write human rules. It is another step to write them on machines that work at all hours regardless of the circumstances.
That’s the gap I keep watching.
Since it is not Fabric that can really stand that pressure, since these digital guardrails do not break when the reality is a bit messy then we are not discussing robots doing the work. What we are dealing with are machines that run within a system that imposes behavior on them just as the laws impose behavior on people.
And in the event that layer becomes real then it is not a question of whether machines can work.
It is who or what makes the rules they follow