Why humanity should not fear AI and you might.

Artificial Intelligence (AI) is a transplant of human intelligence onto a new carrier. It is interesting to observe that such transplant of human intelligence is not new.

The combined work-procedures of large organizations mimic human intelligence. This process can be understood and there is a similarity between the resulting behavior of such organizations and AI systems.

Observing the interaction between organizations and society may be a way to anticipate the interaction with advanced AI systems.

Once the process of such externalization of human intelligence has been observed in organizations, we can see that this process has been going on since long.

The externalized human intelligence carried by organizations should be considered an integral part of humanity. AI systems should also be considered part of humankind, and will augment the chance of survival of humankind. However, part of the biological human beings will degenerate while another part will further develop in cooperation with AI.

Externalization of human intelligence in work-procedures.

Detailed work-procedures are requested or even imposed by democracy, justice, teamwork, business-processes, safety rules, scientific procedures, design procedures, administrative procedures, etc., etc.

These work-procedures are externalized human knowledge describing how an intelligent human being would act under specific circumstances while knowing the background of the job in all its aspects. Work-procedures can be represented in flowcharts as a sequence of activities with interleaving decision points. The more detailed work-procedures get, the more they look like computer software. However they are basically not intended to run on computers, but to be interpreted by human beings.

When a work-procedure is interpreted by a human being, the knowledge embedded in the procedure is exhibited as intelligent behavior of the interpreter, even when this person does not know the reason for this sequence of actions.

The work-procedures play the role of software, while the human interpreter plays the role of processor (computer hardware).

More information about externalization can be found in the references at the end.

Combined work-procedures of an organization.

In a large organization, thousands of humans interpret their detailed job-descriptions in their daily work as described above. The organization as a whole acts like a big machine exhibiting human intelligence originating from the thousands of interlinked job-descriptions. When replacing employees in the organization, the job descriptions are passed on, and the change does not affect the behavior of the organization.

The output of the organization depends far more on the internal structure and the job-descriptions, the work-procedures and the data from earlier activities than on the individuality of the employees.

Even the management cannot change much in that, see for example how small the change is when the government of a country changes completely at once. The changes are even less in organizations where only one person at a time is replaced.

This effect is dramatically amplified by partial-automation of the work-procedures. Remark that the behavior of organizations described above, has always been there, even without computers.

Shift from individual decisions to group-decisions.

Teamwork, democracy, justice, administration, mass-production, globalization, scientific research, all impose to work with increasingly detailed procedures.

Objectivity is aimed at all these levels. Objectivity is the pursuit of decision-making that depend solely on the input and the procedures and is not influenced by the agent executing the procedures, human or machine.

As a consequence every serious decision is made by organizations. Individual decisions become scarce. Therefore, we can safely say that the world is already governed by externalized human intelligence, that the world today is governed by the intelligence of organizations, not by the intelligence of living human beings.

One may object that the intelligence of organizations is also derived from human beings. That is indeed the case, but most of the humans at the origin of the accumulated procedures may be dead since long.

Should we fight that...?
No, no, we fight for the opposite, we fight for democracy, openness in decisions, objectivity at all levels and team-spirit. team-spirit !

So, why fear the upcoming AI while not fearing teamwork and democracy?

BTW: Decisions to invest more or less in AI will be made by organizations, using the intelligence stored in themselves rather than the intelligence of their employees at that moment.

Implications for Humankind.

First, we have to adapt our definition of humankind to what we know. We have to choose to consider the externalized intelligence as part of humanity or not.

If we choose not to include it, the cleverness of humankind drops by 99 %. If we had to divorce, getting the individual human beings at one side, and all organizations, democracy, work-methods, justice-procedures, science, states, and all grouping with procedures at the other side, the biological part of humankind would be very bad-off.

Therefore I assume that we gladly consider the externalized intelligence as part of our humanity, and treat it that way.

If we would have the choice (which is not the case) in making hard-AI or not, we could compare the anticipated future with and without AI to balance the decision.

Future without AI.

In any case the externalization of intelligence will go on. Everything gets more organized, more democratic and more precisely defined. Computerization goes on accelerating. Many parts of the work-procedures get automated, making the organizations gradually more independent from the workers. Intellectual challenges of the work is drastically reduced. The amount of work for human beings drops to a fraction. Human beings gradually loose their intellect. At some point the organizations become rigid and cannot adapt to the world situation, they start breaking down. No humans can interfere by lack of knowledge and intelligence, culture drops back and an important part of the human development has to be redone.

Future with AI.

With AI, at first, about the same thing will happen with the difference that the transitions will go faster and that the organizations will be more human in the sense that they are more aware of their existence in the world and aware of their origins.

Also, because the AI will be integrated in the organizations, they will be far more adaptable. That is because the AI will connect the rootless-knowledge of the work-procedures with the insight of why these procedures are what they are.

Therefore the organizations will take the initiative to form better versions of themselves, which is something that requires an intelligent overview that is beyond the capacity of the human mind.

The insight of AI in the human mind will be used to provide the opportunity for perfect personalized education, lifting up human intelligence of those who accept it.

A separation in humankind between those cooperating with AI and those rejecting it seems inevitable. That is a separation which is already going on today which cannot be stopped since we do not want to restrict the freedom of those who reject technology.

A major advantage of AI is that the chance of long term survival of humankind drastically increases for two reasons.
More intelligent organizations can make better global decisions.
Simultaneously, proliferation of AI in space will start almost immediately, safeguarding the inheritance of humanity from global disasters and protecting itself against sabotage from some groups of humans.

This will start as soon both AI and nanotechnology are ready for it. The information about the biological human and other species will be carried with it, but not in deployed form.

Further reading.

Created Mon Feb 2 19:24:57 CET 2015
Updated Mon Feb 9 16:10 2015
By Luc Claeys.
Send comments to lcl at this site.