Why Build Artificial Intelligence?

Modern life has different kinds of challenges: sustaining a complex civilization, sustaining a high standard of living, and growing, both as a society and as individuals. The ancient Greeks believed that knowledge is gained from pure reason, Renaissance science required experiements to understand the world, and the technological revolution relies on automation.

As digital tools mature, they cease to be single-use assistants for communication, spell checking and record keeping, and increasingly become key in decision making. Simple questions are answered by finding what others have published on the internet, but the era of the search engine is coming to an end.

Complex tasks that so far required the work of teams of specialists can now be automated too. Take for example the mechanical task of programming the path of a plastic extruder (3D printing): the software required is free and readily available to anyone who wishes to use it.

From Games to Science

We now take algorithms that defeat the best chess players for granted, and multi-use algorithms can convert speech to instructions in any home (Alexa). The maid and butler for the upper class has been replaced by the washing machine and Alexa, who can take care of the homes without adding work to humans. This leads to an increase in the standard of living for all, just like automating agriculture has led to a plentiful food supply.

Researchers have so far addressed the purely intellectual problem of “What Can Be Automated?”, and engineers in businesses have converted the concepts into practical machines. This sustains out civilization, increases the standard of living of the individual, and grows the portion of population able to devote energy to higher goals.

But what are the higher goals? Why has humanity done this? Do we wish to rest and loiter and wait for our demise? It would seem the great civilizations of the past have fallen to this, when reaching comfort, they became complacent, and stopped trying. This certainly led to the creation of incredible literary and visual art, but also the downfall of the great Roman empire.

Thanks to science, medicine, and pure logic, we have an endless supply of interesting problems to solve. It makes sense to create machines to improve our health, and further our understanding of the universe. Better medical scanners produce higher resolution imagery, and smarter medical algorithms improve early cancer detection rates. It makes sense to use every available technology to better measure the underlying reality of the universe, so that we may improve our models of physics. But do we want to automate reasoning?

Reasoning Machines

There are two reasons for creating machines that emulate thought: automating work, and extending our intellectual reach. The production of statistical tables used to require years of manual calculation, and can now be computed on-the-fly in a computer. The same breadth-first approach can now be taken in other fields: automatic hypothesis testing, automatic exploration, automatic proof checking. The purpose of this all is to let a single person do the work of a hundred, thereby increasing the pace at which knowledge of reality is gained and converted into useful form.

It will be the work on scientists to improve machines to perform science, and the work of engineers to improve machines that do man’s bidding.

Further Afield

The goals are:

reduction of the load on people doing mundane tasks
increase in capacity for fundamental exploration
the creation of a Von Neumann probe, to colonize the galaxy

### Automating Mundane Tasks

Our global economy relies on people working hard to maintain constant growth. Ownership of the means of production will be critical in maintaining access to a high standard of living, and this is the subject to as many utopias as distopias. Optimism is not enough, access to resources produced by machines will have to become a basic right.

Automating everything from mining to chip manufacture will create an endless supply of tools for our physical and mental well-being.

Fundamental Exploration

In every field of study the tools should be made available to enable their unlimited use: supercomputer access, robotic laboratories, databases of historic records, artistic tools, and others. This applies to every field humanity and individuals wish to explore, both in science and humanities.

Unlike the washing machine, the goal is not to automate the scientist to the point of making him obsolete. Instead, she should strive to have the automatons to increase productivity, and use them effectively. The scientist’s task will be to direct the machines toward goals that are useful or pertinent, without the limitation of resources.

Colonizing Space

When this system will work well on Earth, the issue will become one of resources and space. But why confine ourselves to the spec of dust all of Nature exists on? The most effective way to increase the amount of good done by humanity will be to increase humanity, putting a billion people on each of a billion planets under a common goal of advancement. And why advance at all? Because we are the only way Nature will colonize the universe. As intelligent conservationists of the Earth’s biotope, terraforming barren rocks around all the stars in the sky is the greatest tree-planting philantropic deed we can do.

Conclusion

Our ancestors automated weavers, combine harvesters, and the printing press, so that we may have plenty of clothing, food, and information. Building Artificial Intelligence will make it possible to remove all slavinsh drudgery from life, allowing us to focus on the higher goal of understanding the universe, and the highest goal of brining the galaxy to life.