Further fall in wages will push a big chunk of the middle class into poverty and loss of lifestyle. In rich countries that cushion the fall through unemployment benefits and various subsidies, already know from experience that it encourages people to pull out of the work force and into indolence, especially if families offer extra help. Such state altruism at taxpayers’ expense risks creating a heavy debt burden for future generations. Once the burden is so shifted, it becomes dormant. The other problem resides in the present: profits are more likely to flow to already wealthy shareholders, who save at high rates and hence contribute marginally to demand across the economy. Economists as always will neither anticipate future crises nor have solutions to problems as they arise. Why? AI machines will outsmart them! The worst hit economy is likely to be China due to its heavy dependency on exports; coming out on top is likely to be the U.S. because of its ability to attract migrating talent.
The upcoming 5G technology will further transform society via the IoT. Future networks will have enhanced ability to steer driverless cars, enable surgeons to perform complex surgeries remotely, etc. With IoT, virtual and augmented reality will reach new heights. To survive, the millennials must hone their ingenuity, perseverance, and determination; break molds, take risks, and be resilient; listen to voices of change and progress. The demographic dividend perceived in some developing countries is really a demographic liability (AI is their nemesis). Any unknown phase transition will always present unanticipated blind spots, widely missed by the pundits providing sage advice,79 which with the benefit of hindsight will prove them wrong! The fault lies in assuming that Homo sapiens are sacrosanct while Nature is preparing for their speciation in an entirely novel way. With speciation, questions of ethics, morality, religion, socio economic structure, etc. will undergo a sea change to the extent that Homo sapiens may become an extinct species. There is a wide gulf between knowing that this can happen and believing that it will not happen. Therefore, assessing AI’s impact on social, cultural and political settings may be irrelevant in the long run.
5.4 Will humanoids rule humans?
The robot is defined by its embedded AI. The AI revolution is the most radical transformation human civilization will experience in the post-industrial era. Robots, singly and in groups, can consistently perform flawlessly at peak levels and can combine peak skills. If any of them produces an invention, then en masse all of them can independently produce the same invention on demand. It is this ability that will permit their adopting a dynamic self-organized form of governance without a central authority. This raises a fundamental question: Can humanoids be prevented from becoming our rulers, ruling us with an even hand, rationally, justly, with the philosophy to each according to his ability topped with equitable charity for the disabled needy subject to available resources of goods and services? There will be no need for humanoids to treat natural humans with deference or even treat human life as sacred but as an eradicable epidemical source of disposable ignorance and an undesirable burden on the earth’s limited resources. The humanoids, a unique fusion of the animate and inanimate will have no need for either Heaven or Earth or the men who believe in them. In a world ruled by humanoids, if there is no place for any God, the question of separation of state and religion will no longer arise. Even then, traditional law must undergo a radical change, as must notions of morality, ethics, etc. As Carl Miller notes:
Traditional legal theory holds that to be culpable of a crime, you need intent, or “malice aforethought”. Where’s the intent in an algorithm, especially a randomizing one …? As activism becomes automated, it raises tricky ethical questions we are ill-prepared to deal with.80
However, questions related to the new world order that humanoids may establish can only be idle speculations since humans are unlikely to be consulted by them. Indeed, all such symbols as flags, citizenship, religious beliefs, etc., around which humans rally to proclaim their solidarity as belonging to a region, religion, etc. to defend customs and traditions may become irrelevant. An algorithmically ruled world is unlikely to accommodate such about-to-become historical human relics. Scientists have recently demonstrated in laboratory experiments that a genetic engineering technique known as “gene drive” that uses a gene-editing tool called CRISPR can rapidly spread a self-destructive genetic modification through a complex species, e.g., malaria spreading mosquitoes.81 The risk of Homo sapiens being wiped out by accident or design by such means is no longer unthinkable.
5.5 Progenies of the millennials
With the millennials’ world undergoing dramatic changes brought about by automation and declining real job opportunities, the big question arises: “What about their children, their future?” Children lack a political constituency. Promulgation of laws for their education, health, safety, food, etc. will not guarantee absence of malnourishment, eradicate abusive child labor, or eliminate rampant sexual and physical abuse. It is not yet clear how their education, health care, and safety can be assured. Unless their future is secured, the very survival of Homo sapiens will be jeopardized.
Inability to learn is a death trap. In an AI dominated world, rote education driven by unhealthy politics is misaligned with present needs; only adaptive self-learning will work. Poverty hurts biological development and thus undermines learning. Adaptive self-learning favors those capable of multi-dimensional learning. Homo sapiens may speciate with this feature genetically coded in them. The new species will be explorers if they are to survive in the changing world of work.82 There have been some extraordinary advances in AI in multiple domains as we have noted above. It is too early to understand where they will all lead to or their impact on humanity and the restructuring of society or even the world order. Investment momentum in AI has rapidly increased since 2014 and will not abate soon.83 But one can expect the number of sustainable players in the marketplace to drastically reduce soon while only a few players with deep pockets consolidate their positions.
6 Data-driven world
A quest of AI researchers is finding ways of answering a query. The first significant step in that direction was Google’s search engine. We already know from Gödel’s incompleteness theorems and Turing’s halting theorem that there can be no general search method by which either AI or human intelligence can find answers to all kinds of queries no matter how powerful an axiomatic system we discover. AI machines excel humans because they can dig much deeper into data than any human can. AI thrives on prodigious amounts of data. AI is about big data, deep data, data mining, deep analytics, deep learning. Sources and types of data are many as are the sensors for acquiring them and means of storing them. Major sources of data are objects or entities that emit electro-magnetic signals, audio signals, and textual matter. Of these, data in audio form is comparatively weak and surprisingly less emphasised even though it is rich, e.g., a bat’s world is centered around its auditory sensors and very little on visual sensors. We have even less data related to taste, smell, tactile, and mental. The patterns they encode are likely to produce many surprises.84 In our daily life, such data play pivotal roles in what we do, how we do, why we do, and when we do. Superior species in the future may use such data more advantageously to make the Homo sapiens, the only surviving species in the genus Homo, extinct. No God of any religion has ever mentioned our departed ancestors in the genus Homo. We are ignorant as to how Homo sapiens suddenly acquired language and through it, rational knowledge.
6.1 Data and what we do with it
Big data. Big data is a repository of data collected, say, by a business in a day. Its specific contents may vary by context, e.g., it may include customer or client names, contact information, transaction information, etc. Big data is an enormous resource in business as an input to AI.
Deep data. An aggregation of big data becomes deep data when, paired with experts in a particular knowledge area, it is segregated into useful data from the rest, say, by removing redundancies and annotating it for further analysis. Here one generally deals with exabytes or more of data.
Data mining. It is the art of analyzing large pre-existing databases and extracting information, often by finding correlations hiding in the data.
Deep analytics. It is a data mining process that extracts and organizes bulk data into data structures suited for algorithmic analyses, the analyses thereof, and presentation of output, generally, for easy human comprehension. The real power of data analytics lies in the algorithms used in analyses since they tie the data to a set of concepts and hence “explain” the knowledge hidden in the data. It means the entire data is equivalently condensed into a program plus a subset of the data it uses as input to generate the rest.
[ 79 ] See, e.g., Crawford & Calo (2016), White House (2016a-c).
[ 80 ] Miller (2017).
[ 81 ] Kyrou, et al. (2018).
[ 82 ] WB (2018).
[ 83 ] CB (20180227).
[ 84 ] Kemp (2017). “Almost 150 years before anyone recorded their ultrasound calls, Lazzaro Spallanzani’s cunning yet gruesome experiments revealed how bats navigate in darkness.”