Give us some insights on the AGI?
Roughly, Artificial general intelligence (AGI) is the intelligence of a machine that can understand or learn any intellectual task that a human being can. At Olbrain, we define AGI as a machine capable of achieving a predefined objective in the best possible way under the given boundary conditions.
How is AGI different from AGI-SR?
AGI is generally considered to be the one having human-level Intelligence. The core objective function of humans has been to ensure the survival of the species as a whole, for that we have developed many capabilities that are needed just to maintain a relationship with fellow humans. AGI is supposed to understand these relationships in order to perform well in the Human World.
While in space the scenario is different, the Space Robots need general intelligence to carry out tasks assigned but they do not need to understand humans or the way the human world functions. So, AGI for space robots seems to be achievable within a decade whereas AGI itself may still be far far away.
How is AGI-SR different from AGI for Industrial Robots?
Technology to be developed for Industrial Robots depends on market demand and commercial feasibility. For Example, Amazon warehouse may accept 90% accuracy in pick-and-place through Robots rest they can take care of using humans if it makes commercial sense to them.
While in space human assistance won’t be available so nothing less than 100% efficiency would be acceptable. The upside is, Space Robots can take more time in carrying out the task as compared to their industrial counterparts, and also they can afford to be more expensive.
Why is Robolization important?
We, at Olbrain, have coined the term ‘Robolization’ which stands for Robotic Colonization of space. Since space is way too harsh and cold for human workers to perform, it seems very unlikely that we by ourselves would ever be able to build significantly big structures and manufacturing capabilities to sustain human colonies. We think it’s only the Space Robots who will be the first to colonize space and prepare the grounds for Humans to set their shop.
In how many possible ways can we go extinct?
Humans, as of now, are well equipped to survive on this planet but still, there is no defense in the event of big rare disasters and threats like Asteroid Impact, Large-Scale Volcanism, Global Nuclear Annihilation, Climate Change, Self-Replicating Nanobots, etc. Moreover, the planet earth by design is not supposed to support life forever. It is estimated that conditions to support life would last for the next 1.5 billion years. Also, the Sun will eventually engulf Earth in roughly 7.5 billion years. So it is imperative to expand the civilization beyond Earth and beyond our Solar System.
In fact, we at Olbrain have this hypothesis that Evolution has developed Neo Cortex in Humans with the sole purpose of equipping us with capabilities to go beyond the planet. Though Neo Cortex is a double-edged sword Evolution had no other option.
Explain to us in layman terms, how you are building the control mechanism for the AGI.
Yes, it’s almost impossible to micro-manage AGI and expect it to perform efficiently. AGI would be autonomous at its core, it will be driven by its core objective function which we will hard-code and of course, it won’t be their own survival. But even then at times, the short-term objectives of AGI will clash with our own objectives. To ensure the safety of humans we need to build some control mechanisms. We, at Olbrain, are working on building emotions and feelings to achieve that aim. We believe emotions and feelings in humans too are actually control-mechanisms, Robots also need the same. But yes the emotions of Robots won’t be like that of Humans.