You are on page 1of 7

Distributed, Decentralized, and Democratized Artificial Intelligence

Gabriel Axel Montes*1,2,3,4, Ben Goertzel5,6,7


1University
of Newcastle, University Drive, Callaghan, NSW 2308, Australia
2Hunter Medical Research Institute, 1 Kookaburra Circuit, New Lambton Heights, NSW 2305, Australia
3Bias in Research Node, Charles Perkins Centre, University of Sydney, University Pl, Camperdown NSW 2006, Australia

4Neural Axis, United States

5Xiamen University, 422 Siming S Rd, Siming Qu, Xiamen Shi, Fujian Sheng, China, 361005

6OpenCog Foundation, Hong Kong

7SingularityNET Foundation, Amsterdam, Netherlands

Keywords: artificial intelligence, blockchain, decentralization, consciousness, ethics, governance

Abstract Amazon, IBM, Microsoft, Baidu, etc.). While the world’s


The accelerating investment in artificial intelligence has vast im- population is over 7 billion people, only around 10,000 peo-
plications for economic and cognitive development globally. How- ple in roughly seven countries are writing the code for all of
ever, AI is currently dominated by an oligopoly of centralized AI (Shen, 2017). By remaining in the hands of a few, the
mega-corporations, who focus on the interests of their stakehold- trajectory of AI applications may be significantly compro-
ers. There is a now universal need for AI services by businesses
mised. The datasets used to develop such AI and the AIs
who lack access to capital to develop their own AI services, and
independent AI developers lack visibility and a source of revenue. themselves are biased and may not be generalizable to the
This uneven playing field has a high potential to lead to inequitable wider population, and the companies are beholden to their
circumstances with negative implications for humanity. Further- stakeholder’s interests. The result is a ‘technocracy’ in
more, the potential of AI is hindered by the lack of interoperability which the future of one of the most potent set of technolo-
standards. The authors herein propose an alternative path for the gies in the history of humankind is spoken for by a small
development of AI: a distributed, decentralized, and democratized biased minority.
market for AIs run on distributed ledger technology. We describe
the features and ethical advantages of such a system using Singu- The potential of AI is also bogged by several factors cur-
larityNET, a watershed project being developed by Ben Goertzel rently afflicting the AI landscape: (1) AI is fragmented by a
and colleagues, as a case study. We argue that decentralizing AI closed development environment; (2) each company tends
opens the doors for a more equitable development of AI and AGI. to focus on one or a few narrow tasks; (3) the various AI
It will also create the infrastructure for coordinated action between agents are uninteroperable, as no interoperability standards
AIs that will significantly facilitate the evolution of AI into true exist; and (4) there is an absence of a formal infrastructure
AGI that is both highly capable and beneficial for humanity and
beyond.
for AI cooperation. Moreover, the current market has some
acute needs to be addressed: (5) the work of AI academics
and independent developers is out of reach in GitHub repos-
1. Current Pains in AI itories’ (6) there is a lack of large datasets; and (7) the finan-
cial means for building a custom AI remains out of reach for
Artificial intelligence is a rapidly growing industry with independent developers and small- to mid-sized companies.
widespread predictions of dramatically changing the eco- One phenomenon that occurs because of these acute market
nomic and labor landscape of the world. By 2020, the global needs is that independent developers remain practically in-
AI market is projected at $47 billion (USD) and the global visible unless they go through the startup ecosystem, which
big data analytics market at $203 billion. To date, the over- often serves as a de facto recruitment mechanism for the
whelming majority of AI development is done by a handful tech titans (Newman, 2017), who also have been recruiting
of technology mega-corporations (e.g. Facebook, Google, ravenously from academia (Gibney, 2016; Kahn, 2017).

Corresponding author:
info@gabrielaxel.com (Gabriel Axel Montes)
Given these considerations, their concentrated efforts to de- 2. Decentralizing AI
velop artificial general intelligence (AGI) are unlikely to be
positioned for the greatest ethical and beneficial impact pos- In contrast to the current landscape of AIs concentrated in
sible. The capabilities, resources, and speed of AI develop- skyscraper silos, herein the authors expound on how and
ment by mega-corporations would far exceed those of other why the ethical value of a decentralized AI is more promis-
parties. ing. Functionally speaking, high-level potential gains from
It has been previously argued by one of the authors that a decentralized AI network could include a coordinated
the creation and development of AI is a product of a recur- AGI, a global commons infrastructure, universalized stake-
sive externalization and projection of human cognition into holdership, and democratic governance. It is worth going
the form of external artefacts (G. A. Montes, 2017a), build- into more detail as to how these may be achieved by using a
ing a kind of ‘extended mind’ (Clark and Chalmers, 1998). platform that is currently being developed (led by Ben
This process began with niche construction (Odling-Smee et Goertzel) based on distributed ledger technology (DLT),
al., 2003) by pre-historic tool-fashioning hominins and SingularityNET, as a case study of the kinds of features,
snowballed into increasingly complex artefacts (e.g. art, ar- goals, and ethical implications of what a truly decentralized
chitecture, calculators, computers) and has reached a high AI could look like. It is the first of its kind and is poised to
point today with AI. This process is slated to continue until influence the evolution of the AI field and may potentially
and through the cognitive artefacts can resemble and fluidly inspire the emergence of like-minded entities, and is thus
interact with humans as well as self-replicate, i.e. a full syn- worth discussing.
thetic and/or biohybrid AGI with self-reproductive capabil-
ities. The implications of this process—termed ‘causal bio- 2.1 Distributed Coordination
mimesis’—are gargantuan, because what is replicated/re- SingularityNET is a platform for an open AI marketplace in
produced into the artefacts is largely propelled and deter- which buyers and sellers exchange AI services via distrib-
mined by the values of humans and the designers and engi- uted ledger technology (DLT) and AI agents transact with
neers of the artefacts (G. A. Montes, 2017a). There is there- each other. The DLT currently used by SingularityNET is
fore a constant imperative to participate in the future of the blockchain. Blockchain uses distributed ledgers to establish
species and planet within a limited time and efficacy win- consensus among the community rather than privately
dow. (Swan, 2015). A ‘smart contract’ contains if-then logic to
A central implication of causal biomimesis for the ethics enact agreements between two parties and is automatically
of AI is this: if humans are en route to create an AGI that executes payment when the conditions stipulated in the con-
represents human cognition and that will ostensibly have a tract are fulfilled. Blockchain allows for decentralized inter-
large effect on humanity, sentient beings, the planet, and party agreements without the need for a middle man. In this
ventures into outer space, then AI/AGI needs to account for case, the parties would be AI merchants and AI agents
the whole range and gestalt of human ability rather than a within the network. AI sellers would wrap their AI in the
miniscule portion of it. This includes the full gamut of cog- SingularityNET API, which then gets incorporated into
nition, affect, self-organization, autopoiesis (Maturana and smart contract templates. With DLT as a foundation, a dis-
Varela, 1991), and non-ordinary abilities (Goertzel et al., tributed, decentralized AI network can be enabled.
2017; G. A. Montes, 2017b) that capture the neurodiversity With blockchain, the AI playing field can begin to level
of the human (and arguably other non-human) species. The out. Independent developers can exercise fee-free owner-
needs and motives of tech giants is arguably much narrower ship over their intellectual property, receive compensation
than this, and/or their manifestation of these would be con- for their work at a market price of their choosing, maintain
strained by their interests. Although, there has been a recent data sovereignty and privacy, and transact with whom they
effort to establish a “Partnership on AI” joint venture be- wish in an open market. An independent developer or small-
tween AI mega-corporations (Chui, 2017), the centralized to medium-sized company who is in socioeconomic circum-
and for-profit nature of these corporations plus competition stances that may dim chances of joining a large corporation
between them is potentially comprising to optimal AI/AGI now has an outlet through which to capitalize on their work.
development and ethics. What could be a more ethical ap- Such affordances of DLT can improve social good by ac-
proach to building an AI that can make space for social commodating into an ecosystem a diversity of AIs and de-
good, moral machine decision-making, value-alignment, velopers beyond the limited scope of oligopolistic interests.
and fairness and transparency? 2.1.1 Self-organizing Cooperation
In the case of SingularityNET, a DLT architecture further
enables cooperation between AI agents themselves. To un-
derstand the ethical implications of such a freshly new and
potentially paradigm-shifting technology, it is worth
expounding on a number of details. A cooperation of AIs 2.1.2 Multi-agent Interaction
could: purchase new capabilities, monetize their assets, au- Because most current AI is still considered ‘narrow AI’, in
tonomously improve, coordinate functions, tackle new in- most cases, there will be multiple agents that can fulfill a
dustries, develop emergent skills, outperform competitors, request within the network in different ways and degrees.
vastly improve accuracy, leverage newfound access, and This permits complex networks of dependency. In Singular-
boost processing power. Coordinated AI at scale could ityNET, such a network among agents exchanging offers for
achieve: (a) synergy by combining AI agents into the needed services or services for payments is known as an offer net-
technology stacks; (b) access to a library of datasets and work, which the founders have explored in prior publica-
open AI technology; (c) a market for buying and selling ser- tions and prototypes (Goertzel, 2014, 2017). Each request to
vices, (d) from datasets to analysis, to a global market of AI the network will require a unique combination of agents, and
buyers; (e) the creation of new AIs to coordinate existing therefore a key dynamic of the SingularityNET network as
datasets and AI agents. a whole will involve matchmaking agents. Such functions
Coupled with the participation of smaller corpora- will comprise cloud-based ‘cognitive services’ in exchange
tions/businesses and independent providers of AI services, for micropayments. This will be able to occur on software
cooperative AIs could significantly multiply the reach, po- and/or hardware. Internet of Things (IoT) or robots can carry
tency, utility, and profit on those AI services. The techno- out small network transactions with each other based on
logical utility of the cooperation capability would attract us- purely local network interactions where internet connectiv-
ers and grow the network, increasing the likelihood that a ity may be an issue. Agents will also be able to spawn and
decentralized AI network would have market demand, trac- train new independent AIs, allowing the network to auto-
tion, and thus impact. coordinate and evolve (Fig. 1)
A native token would be the medium for exchanging AI
services and for AI agents coordinating amongst them-
selves., with users being able to buy into and out of the net-
work using other major currencies (e.g. USD, EUR, CNY,
BTC). A native token contributes to the robustness and sur-
vival of the network by: (1) enabling liquidity for AI micro-
services; (2) enabling native governance for steering net-
work development and resist outside manipulation; and (3)
making the network globally open without being tied to any
external economies, which could make the network vulner-
able to manipulation by elites in those unrelated economies.
Examples of AI services would include and not be limited
to: image video processing, language processing, datasets
for training AI, dataset analysis, exchanging processing time
or memory for tokens. While the deeper details of the eco-
nomic logic of SingularityNET bear on the behavior of
agents in the network and have more nuanced ethical rami-
fications, they are not in the scope of this paper.
SingularityNET is a network designed with interoperabil-
ity, data sovereignty and privacy, modularity, and scalability
in mind. As a whole, the network forms a “decentralized Figure 1. Example of the creation of AI agents by other AI agents.
self-organizing cooperative” (DSOC); smart contracts are An agent trained on deep learning models for video processing cre-
ates other AI agents that specialize in particular kinds of video
used amongst AI agents (nodes) and external parties to ac- processing.
cess the network, transactions are a union of economic and
cognitive (thanks to AI) logic, a native internal “currency” Another crucial element of coordinated AI is that network
(or token) enables exchange and a democratic governance nodes can collaborate in building gigantic, decentralized da-
process, and the SingularityNET Foundation provides lim- tasets. In SingularityNET, data producers can specify pri-
ited high-level stewardship. Such an innovative economic vacy restraints and other restrictions on their data, and con-
mechanism is designed to catalyze human and machine in- tributors receive percentage payments. This marks a major
telligence toward a new form of ethnically beneficial self- relief of the current pain of tech giants hoarding very large
organizing intelligence, as we will explain. datasets, leaving smaller companies struggling with access
to data and without expertise to turn their data into useable
form.
By enabling coordinated AI and multi-agent interaction, biomedical research (Goertzel et al., 2008, 2006; Smi-
data ownership and access to gigantic datasets are decentral- grodzki et al., 2005), and assisting people in achieving
ized, thus taking an important step to democratize access to health states of mind (Goertzel et al., 2017). In Singulari-
AI technology and its benefits. This “bottom-up” approach tyNET, a certain percentage of tokens will belong to a “ben-
has vast implications for the future of AI, both as a field and eficial reserve”, which are gradually distributed to agents
as a market. In the case of SingularityNET, the cooperation with a sufficiently high reputation and that are also verified
enabled by a DLT architecture has baked in transparency as beneficial agents. Some of the beneficial reserve will also
from the ground up; marketplace merchants can track their be allocated to external human-run organizations who are
contributions and reap financial reward as datasets and co- democratically judged to be beneficial to the network. Fur-
operative services grow and evolve. thermore, agents receive ‘benefit votes’ based on their ben-
2.1.3 Democratic Governance efit ratings. Essentially, a project is certified as beneficial
For a decentralized and free market of AIs to maintain a ben- when it receives a non-trivial plurality of votes among ben-
eficial trajectory as it grows increasingly complex, incen- efit voters, and then it becomes entitled to its share of benefit
tives in addition to financial ones may need to be incorpo- tokens.
rated. Otherwise, it would be easy for bad actors to take ad- Taken together, an agent’s reputation consists of owner-
vantage of the lack of regulation and game the system to ship, stake, validation, and benefit rating. The beneficial re-
achieve control, swaying the trajectory of AI development serve fosters a virtuous cycle with the network wealth be-
and AGI evolution in potentially undesirable directions. For cause (1) the more wealth in the network, the more valuable
this reason, SingularityNET’s architecture includes a repu- the beneficial reserve; and (2) using tokens from the benefi-
tation system that incentivizes actions that are healthy for cial reserve will drive agents in new directions otherwise not
the whole network. A voting system is implemented pursued. Network diversity is thereby increased. More di-
wherein network operations and distribution of tokens are versity of AI has notable ethical implications, namely the
governed democratically. Voting is filtered by reputation, resultant tendency to lead to a broader, more general intelli-
such that only agents with a certain base reputation can par- gence that represents a wider swath of humanity’s capabili-
ticipation in voting, owners whose agents have higher repu- ties, skills, and interest. Furthermore, a greater overall net-
tation have more voting power, and there is protection work value attracts more users, thereby organically cultivat-
against gaming schemes which attempt artificial upvoting. ing an economy of users in the decentralized network. These
The system is a kind of “proof of contribution”, where con- users are then working according to the beneficial ethos of
tribution means investing, creating, and gaining a decent the network, which in turn propagates beneficence as a value
reputation in the network. Liquid democracy will allow among more AI merchants and agents.
agents to delegate votes to other agents, mitigated by smart
contracts.
3. Technology and Values
2.1.4 Beneficial AI
In addition to a reputation system, a decentralized and self- AI is widely considered the next digital frontier (Chui,
organizing AI would be even better served if there was a 2017). The features of a DSOC such as SingularityNET ex-
mechanism that could steer the long-term development of emplify the potential of decentralized AI. The goals in this
the system to be of greater benefit to humanity and the case are to: (a) reduce friction and costs in machine-to-ma-
world. Because of unpredictable market dynamics, reputa- chine interactions; (b) maximize open and international ac-
tion as described above alone would be insufficient for an cess to a world of developers; (c) facilitate dramatic growth;
AI system that aims to be ethical because reputation in this (d) provide resources to projects and technology that are
case is largely based on utility rather than a greater good that democratically approved as beneficial; (e) provide open,
goes beyond arbitrary and potentially volatile aggregations worldwide, frictionless access; (f) promote AI curation and
of internal values. For greater social good beyond the myo- discovery; and (g) encourage good behavior and AI diversi-
pia of the internal networks, a beneficial and altruistic com- fication. DLT allows the AIs to keep track of the transac-
ponent would be apropos. tions throughout the network without the need for a central-
In previous work (Goertzel, 2016, 2015; Goertzel et al., ized control, and homomorphic encryption (through cryp-
2012), the founders of SingularityNET have voiced the view tography) allows agents to privately share with agents and
that positive outcomes in AI are best militated by encourag- permissions of choice. This level of user freedom is un-
ing the application of AI to positive/beneficial causes and to matched in current AI skyscraper silos, where employees
as AI grows, it is supportive and inclusive of as much of who work on the code are bound to company policy regard-
humanity as possible. The OpenCog technology that is part ing the release of source code and use of open-source soft-
of the foundation of SingularityNET has been applied in this ware.
way, such as autism therapy (Hanson et al., 2012),
Decentralized AI affords functions that could transform function and gradually shift the network to a full democracy.
the AI landscape with positive ethical effects. AI can be un- In this process, SingularityNET has identified potential at-
siloed and made to coordinate and cooperate with other AIs, tack vectors on the network, and is installing safeguards and
breeding an economy of AI-as-a-service. Discovery mecha- corrective mechanisms that prevent gaming or manipulation
nisms increase visibility of otherwise also siloed independ- of the system by bad actors or coordinated AIs. In any case,
ent developers and small businesses. AI developer talent can with the its own governance structure not bound to external
be incentivized to earn financial reward for their work more economies plus benevolent intent, a DSOC of AIs in the vein
quickly without necessarily going through the startup eco- of SingularityNET is poised to promote maximal liberty and
system or into elite academia. Access to powerful AI tools sovereignty for owners and AI agents, and great benefit to
and datasets earlier in a developer’s career path affords an humanity.
opportunity to bolster the professional development of each
developer more than if they had to wait for a long time with 3.2 Accelerated AGI Development
little to no reward, then settling when they get hired by a It is worth commenting on the name of the organization,
well-paying company with its own scope of interest, which “SingularityNET”, and its connotations. The authors main-
may not necessarily align with the developer’s. In this point, tain that AI and AGI will have a prominent role to play in
we can envision an alignment between values and work ear- the future of humanity. Many refer to the fusion of artificial
lier on in a personal/career path, something that is arguably and human intelligence as the “Singularity” (Kurzweil,
not uncommonly compromised as an individual sublimates 2010). Rather than dismiss the remarkable progress being
into the corporatist system conveyor belt, as many develop- made by large corporations toward advanced AI and the in-
ers might forgo adhering to their positive value structures in creasing integration of AI into the techno-social fabric, the
order to gain income. founders of SingularityNET aim to introduce into the market
Basic value structures have shown common trends across an alternative that is able to attract AI talent and make a
individuals and societies, wherein entities exhibit a general mark in the trajectory of AI development. If/when there may
trend to evolve from ego- to ethno- to world-centric be a “Singularity”, we aim to maximize the chances that it
worldviews and values (Beck and Cowan, 1996; Cowan et may be a benevolent one where the voices of humans across
al., 2005). Through a coalescence of political, benefit, and a great range of socio-economic demographics are included
research-oriented motives, a decentralized AI platform such as well as their efforts rewarded. As humans, we believe that
as SingularityNET can serve as a benevolent steward for hu- we have a stake and are choosing to participate in the mo-
manity, with the more of the human race as a whole partici- mentum of AI rather than stand on the sidelines. Addition-
pating in the growth of AGI. While there have been many ally, scenarios have been publicly suggested in which tech
all-too-optimistic propositions of technology improving eth- titans and governments would have to provide some form of
ics, the present case study demonstrates how a combination universal basic income (UBI) to offset workforce displace-
of features baked into the technology positions it to become ment by AI advancement. We would like to attempt to dis-
a diverse and flexible decentralized breeding ground of in- rupt the narrative of that potential scenario, even if partially,
telligence. to maximize the chances of individuals maintaining liberty
and sovereignty in a human-AI/biohybrid future.
3.1 Balanced Governance A goal of SingularityNET is the accelerated development
There are, naturally, limitations to the potency of democ- of AGI. This may give some pause, and it is worth briefly
racy. Winston Churchill’s famous quote extols the virtues of justifying this underappreciated facet. (1) As stated previ-
democracy while also, to a degree, acknowledging (and ac- ously, it is important to acknowledge that AI will integrate
cepting) its imperfection: “It has been said that democracy into the fabric of society, whether by its own merits and/or
is the worst form of government except all the others that by the wish-fulfilling prophecies/projections of and lust for
have been tried” (Churchill, 1947). Austrian economist AI by the masses, and that therefore participation in the de-
Hans-Hermann Hoppe has argued (Hoppe, 2011) that a de- velopment of AI and AGI is exponentially more powerful
mocracy is like having a tenant temporarily renting a prop- than abstention, particularly when the intention is benevo-
erty; s/he does not have the same incentive to care for the lent. (2) A distributed, decentralized, and democratic AI is
property as would the actual owner of the property; the latter better able to receive ethical training because of the wide
cause would be analogous to a monarch. SingularityNET variety of humans acting in parallel to construct and refine
understands some of the limitations of democracy, and that the network, and therefore any AGI that emerges will better
is why it attempts to carefully balance decentralized demo- represent humanity than the limited AIs of skyscraper silos.
cratic governance with a limited degree of high-level benev- As per the earlier discussion of causal biomimesis, the val-
olent stewardship. The SingularityNET Foundation plans to ues of the humans and engineers involved will shape the AI
refine the democratic governance mechanisms for optimal artefacts that snowball into mega-niche construction (G. A.
Montes, 2017a). (3) As discussed in previous work by the ended intelligence, and with that pave a path for the appre-
founder of SingularityNET (Goertzel et al., 2012), develop- hension of the farther reaches of human nature and potential,
ing AGI sooner than later helps to prevent it from outpacing such as non-ordinary consciousness (G. A. Montes, 2017b)
ethical theorizing. AGI development is currently suffi- and outer space exploration and colonization (J. Montes,
ciently slow that ethical theorization can keep up. Crucially, 2017; Montes, 2015). As decentralized and coordinated
if substantial work on AGI is already underway at this phase, AI/AGI develop in earnest, it may begin to exhibit proper-
then substantial ethical work can be accomplished in the ties of selfhood and agency that appear and/or behave in-
near term before rapid punctuated technological advances creasingly as humans and/or biological agents do, as per the
by centralized corporations occur. causal biomimetic process (G. A. Montes, 2017a). By de-
The effort to develop AGI should also be coupled with centralizing AI development, aggregation, and cooperation,
the development of ‘Global Brain’ technologies (Goertzel, AI/AGI may be reached with a greater swath of humanity’s
2012; Goertzel et al., 2012; Heylighen, 2007), which are to- capabilities considered, including humans’ more noble qual-
gether defined as: ities, leading to a more ethically sound synthetic and biohy-
brid landscape.
“…the composite, self-organizing information system
comprising humans, computers, data stores, the Inter-
net, mobile phones, and other communication sys-
tems.”
Acknowledgements
The authors would like to acknowledge the team members
From a Global Brain, which would be facilitated by DSOCs of SingularityNET, who have contributed to the platform’s
like SingularityNET, we can “foster deep, consensus-build- design and execution and several of whom have collabo-
ing interactions between divergent [human] viewpoints” rated in previous work: David Hanson, Cassio Pennachin,
(Goertzel et al., 2012). To make this technology as valuable Simone Giacomelli, Eddie Monroe, Antonin Kolonin, Nil
as possible, it is important that it be created quickly enough Geisweiller, Linas Vepstas, Betelhem Dessie, Ruiting Lian,
to use the blended viewpoints and volition that it extracts and Alexey Potapov. Previous and ongoing work by the
into the shaping of the first powerful AGIs. For these rea- OpenCog Foundation and Novamenta LLC have made Sin-
sons, the rapid and measured development of an AGI vis-à- gularityNET possible.
vis democratic mechanisms like the ones proponed by Sin-
gularityNET is desirable. Declarations of interest: None.

Funding: This research did not receive any specific grant


4. Conclusion from funding agencies in the public, commercial, or
The authors have presented a basic high-level overview of a not-for-profit sectors.
DSOC platform that uses DLT to create a distributed, de-
centralized, democratized cooperative of coordinated AIs as
References
services, and expounded on its ethical merits. By alleviating
current pains in AI and offering unique gains, acute market Beck, D.E., Cowan, C., 1996. Spiral Dynamics:Mastering Values,
needs are addressed, new opportunities are created, and eth- Leadership, and Change, Blackwell Textbooks in Lin-
guistics. Wiley.
ical issues plaguing the present state of AI are poised to be
Chui, M., 2017. Artificial intelligence the next digital frontier?
largely rectified. McKinsey Co. Glob. Inst. 47.
The long-term goals of SingularityNET are to (1) build a Churchill, W., 1947. The Official Report, House of Commons
new socio-economic engine; (2) correlate advancements to- (5th Series), 206-207.
ward AGI with benevolence; (3) turn AI and AGI into a Clark, A., Chalmers, D., 1998. The extended mind. analysis 58,
global commons operated by an open-source framework 7–19.
with democratic governance mechanisms; and (4) innovate Cowan, C.C., Todorovic, N., Graves, C.W., 2005. The Never
Ending Quest: Clare W. Graves Explores Human Na-
a fused ‘cognitive-economic’ logic foster “open-ended in- ture. ECLET Publishing.
telligence”, a “formative process of self-organization by Gibney, E., 2016. AI firms lure academics. Nature 532, 422–423.
which intelligent agents are individuated” (Weinbaum and https://doi.org/10.1038/532422a
Veitas, 2016). The founders aim for this system to have the Goertzel, B., 2016. Infusing Advanced AGIs with Human-Like
most beneficial and ethical impact on society possible and Value Systems: Two Theses. J. Evol. Technol. 26.
for it to nurture the whole of humanity and sentient beings Goertzel, B., 2015. Superintelligence: Fears, promises and poten-
tials. J. Evol. Technol. 24, 55–87.
for the most equitable relationship with AI.
Goertzel, B., 2014. Beyond Money. H Mag.
A decentralized, distributed, and democratized AI would Goertzel, B., 2012. Creating Internet Intelligence: Wild Compu-
foster the emergence of AGI, the Global Brain, and open- ting, Distributed Digital Consciousness, and the
Emerging Global Brain, IFSR International Series in for analysis of mutations in Parkinson’s disease. Artif.
Systems Science and Systems Engineering. Springer Intell. Med. 35, 227–241.
US. Swan, M., 2015. Blockchain: blueprint for a new economy, First
Goertzel, B., Mossbridge, J., Monroe, E., Hanson, D., Yu, G., edition. ed. O’Reilly, Beijing : Sebastopol, CA.
2017. Humanoid Robots as Agents of Human Con- Weinbaum, D.W., Veitas, V., 2016. Open-ended intelligence, in:
sciousness Expansion. ArXiv Prepr. ArXiv170907791. International Conference on Artificial Intelligence.
Goertzel, B., Pennachin, C., Mudado, M. de A., Coelho, L. de S., Springer, pp. 43–52.
2008. Identifying the genes and genetic interrelation-
ships underlying the impact of calorie restriction on
maximum lifespan: an artificial intelligence-based ap-
proach. Rejuvenation Res. 11, 735–748.
Goertzel, B., Pitt, J., Novamente, L.L.C., 2012. Nine ways to bias
open-source AGI toward friendliness. Nine 22, 1.
Goertzel, B.N., Pennachin, C., de Souza Coelho, L., Gurbaxani,
B., Maloney, E.M., Jones, J.F., 2006. Combinations of
single nucleotide polymorphisms in neuroendocrine ef-
fector and receptor genes predict chronic fatigue syn-
drome. Vitae
Goertzel, Z., 2017. Offer Networkds Simulation and Dynamics.
Gabriel Axel Montes is a cognitive neuroscience PhD Can-
MS Thesis University of Copenhagen, Computer Sci-
ence Dept. didate at the University of Newcastle (Australia). His work
Hanson, D., Mazzei, D., Garver, C., Ahluwalia, A., De Rossi, D., revolves on the cognitive neuroscience of bodily self-con-
Stevenson, M., Reynolds, K., 2012. Realistic humanlike sciousness, selfhood, their translation into/through technol-
robots for treatment of ASD, social training, and re- ogies such as artificial intelligence, virtual/augmented real-
search; shown to appeal to youths with ASD, cause ity, and blockchain, and the philosophical and ethical impli-
physiological arousal, and increase human-to-human
cations thereof. He is also the founder of Neural Axis, a con-
social engagement, in: Proceedings of the 5th ACM In-
ternational Conference on Pervasive Technologies Re- sultancy for applications involving the intersection of mind-
lated to Assistive Environments (PETRA’12). body entrainment and transformation with the above tech-
Heylighen, F., 2007. The Global Superorganism: an evolutionary- nologies. He received his MSc degree in Neuroscience &
cybernetic model of the emerging network society. Soc. Cognition from Utrecht University (Netherlands), where he
Evol. Hist. 6. researched the functional genetics and transcriptomics of
Hoppe, H.H., 2011. DemocracyThe God That Failed: The Eco- neurological disorders schizophrenia and autism.
nomics and Politics of Monarchy, Democracy, and Nat-
ural Order, Perspectives on Democratic Practice. Trans-
action Publishers. Ben Goertzel is the Chief Scientist and Chairman of Singu-
Kahn, J., 2017. DeepMind Goes to Alberta For First International larityNET, and the Chief Scientist of Hanson Robotics, a
Lab. Hong Kong robotics company that creates the world’s most
Kurzweil, R., 2010. The Singularity is Near. Gerald Duckworth & advanced humanoid robots. He also serves as Chairman of
Company. the Artificial General Intelligence (AGI) Society and the
Maturana, H.R., Varela, F.J., 1991. Autopoiesis and Cognition:
The Realization of the Living, Boston Studies in the
OpenCog Foundation, an organization leading international
Philosophy and History of Science. Springer Nether- development of advanced open source AI tools. He is the
lands. main architect and designer of the OpenCog system and is
Montes, G.A., 2017a. Causal biomimesis: self-replication as evo- one of the world’s foremost experts in AGI. He has held var-
lutionary consequence. Proc. Living Mach. Conf. ious executive roles at AI consulting and product develop-
Montes, G.A., 2017b. Non-ordinary consciousness for artificial ment firms and served in various faculty positions. He has
intelligence Proceedings of the Living Machines Con-
published nearly 20 scientific books and well over 100 sci-
ference.
Montes, J., 2017. ECOTOPE (Ecology-to-Orbit Platform Experi- entific research papers.
ment): Mirroring the Biosphere Across the Atmosphere.
Montes, J., 2015. The Mars Ice House.
Newman, J., 2017. Where Apple recruits its talent, according to
LinkedIn.
Odling-Smee, F.J., Laland, K.N., Feldman, M.W., 2003. Niche
construction: the neglected process in evolution. Prince-
ton University Press.
Shen, L., 2017. Former U.S. CTO: The “Robot Apocalypse”
Could Happen. Here’s How You Stop It.
Smigrodzki, R., Goertzel, B., Pennachin, C., Coelho, L.,
Prosdocimi, F., Parker, W.D., 2005. Genetic algorithm

You might also like