The race is on to complete the global nervous system of civilization and make supercomputing power and artificial intelligence available to everyone. Humanity, the built environment, and ubiquitous computing are becoming a continuum of human consciousness and advanced technology reflecting the full range of human behavior, from individual philanthropy to organized crime. Elon Musk’s Starlink is completing global access to the Internet with an advanced broadband system via over 5,500 micro-satellites today with plans for 42,000 micro-satellites; China plans a similar system with 26,000 satellites; and Amazon’s Project Kuiper plans its system with 3,200 micro-satellites. The Internet connects 5.35 billion users as of April 2024 and is expected to connect nearly everyone within ten years. Assuming Artificial General Intelligence could be available in the same time period, then the majority of humanity would have access to global intelligence that gets smarter continuously. Today, two-thirds of the world has a mobile phone; over half have smart phones, which is forecast to be 7.33 billion by 2025. The continued development and proliferation of smart phone apps are putting state-of-the-art artificial narrow intelligence systems in the palm of many hands around the world. The USA’s Frontier is the fastest computer with 1.1 Exaflops / second, followed by Japan’s Fugaku with 442 petaflops and Finland’s LUMI with 152 petaflops and then USA’s Summit and Sierra both from IBM, and Sunway TaihuLight from China. The race is also on for large-scale quantum computing, that promises to solve vastly more difficult and complex problems, far faster than super computers within five to ten years. IBM’s Condor is leading in 2024 with 1,121 superconducting qubits. Some examples of less powerful quantum computers, but commercially available today are USA’s IBM quantum, QuEra, and IonQ Forte, Canada’s D-Wave, China’s Origin Wukong, and UK’s Rigetti.
Another great race today is to develop artificial general intelligence, which some think could arrive within five years. If so, its impact will be far beyond current artificial narrow intelligence. However, if we don’t get the initial conditions, guardrails, regulations, and governance “right” for both the AGI developers and users, then many unregulated AGIs could evolve into an artificial superintelligence beyond our awareness that science fiction has warned about. Investments in AGI are forecast to reach $200 billion globally by 2025 with a market volume of $826.70 billion by 2030. To develop safe AGI, OpenAI, Microsoft, Google, and Anthropic have created the Frontier Model Forum, and IBM, Meta, and 50 others such as Intel, NASA, CERN, and the University of Tokyo have created the AI Alliance. Representatives of the U.S. and China have been secretly meeting in Geneva since July 2023 to coordinate safety strategies. The U.S., China, and 26 other countries sign the Bletchley Declaration to collaborate on future AI safety. The UN General Assembly adopted a U.S. initiated AI resolution on safety and China initiated a second UN resolution on collaboration on AI development also adopted by the UN General Assembly. China has declared its goal of being the world’s AI leader by 2030. President Putin of Russia said whoever leads in AI rules the world. International and national AI strategies are being continually updated as the race for AGI continues.
Meanwhile, DeepMind’s Gato (of Alphabet) is beginning the transition from ANI to AGI with over 600 different tasks from the control of a robot to recognizing objects. The human brain projects of U.S., EU, China, and other countries, plus corporate ANI and AGI research, should lead to augmented individual human and collective intelligence.
Web 3.0 promises to integrate immersive augmented reality and virtual reality metaverse with decentralized user-centric systems like blockchain, Internet of Things (IoT), decentralized autonomous organizations, a variety of artificial narrow intelligences (from autonomous agents to natural language processing), and semantic web technologies, all integrating our physical and digital lives. Web 3.0 will evolve in parallel with Web 2.0. However, such decentralized systems could make enforcement of regulations against disinformation and information warfare more difficult and give vastly more entry points for cybercrimes. As global tele-smart-everything evolves, how well governments develop and coordinate Internet security technology and regulations may determine the future quality of cyberspace.
Cybercrime increased by 600% during the Covid pandemic. Ransomware accounts for 70% of all malware payloads. The global cost of cybercrime is estimated to grow to $10.5 trillion annually by 2025. With the evolution of the Internet of Things, wearable computers, autonomous vehicles, and brain-computer interfaces, cyber security will become increasingly important. Low-cost computers are replacing high-cost weapons as an instrument of power in asymmetrical cyber and information warfare. Information security should address a wide and diverse range of enemies — from the “geek in the back room” to criminal organizations and governments. To prevent cheating on exams and suppress political dissent, 35 governments shut down the Internet 187 times, slightly up from 34 countries that shut down the Internet 182 times in 2022, and compared to 25 countries that shut down the Internet 213 times in 2019. In 2023, 18 governments intentionally interrupted connectivity 124 times including 55 nationwide disruptions totaling 2370 days of disruptions.
As of April 2024, 64.6% of the world was connected to the Internet (5.44 billion users) and 5.74 billion use social media, which is expected to increase to nearly six billion by 2027. Social media users averaged 151 minutes per day in 2023, up from 147 minutes from the previous year. Social media has connected people to improve their lives, but it also intensifies polarization by feeding users more extreme versions of their preferences to compete for user’s attention and clicks for advertisers. As a result, social media has become a powerful tool used in information warfare. Facebook reported that it deleted 1.8 billion fake accounts in the third quarter of 2021. While the numbers have come down since then, another 827 million fake accounts were deleted in the third quarter of 2023.
Internet advertising income passed television ad revenue as long ago as 2017 and continues to grow, with mobile phones now accounting for half of all internet traffic. Rapidly increasing video, AR / VR, and IoT use raises concerns about anticipating and meeting future bandwidth demands for an internet infrastructure that was not designed for these applications. Internet reliability has become strategically vital for much of civilization. More than a billion hours are viewed each day on YouTube.
Blockchain is being explored as a new approach to IoT security, as is quantum entanglement. Quantum cryptography is an emerging security technology in which two parties can generate shared, secret cryptographic material between ground stations or between Earth and satellite (as demonstrated in June 2017 by China). Norton has seen a 51% increase in password stealing malware during 2022, with over 24 billion passwords exposed by hackers in 2022 alone. Future quantum computers could make current security systems obsolete; NIST is working to prevent this with post-quantum security algorithms and standards. In December 2020 China announced a quantum computer calculation in 200 seconds at room temperature that a supercomputer would take 2.5 billion years to complete. IBM’s 127 qubit chip is available for commercial use and IBM has produced a 2-nanometer chip with 50 billion transistors on the size of a fingernail increasing performance and reducing energy demand. ASML also pushes Moore’s law with its chips at 3-7 nanometers. Both IBM and Google announced they plan to have a 1-million-qubit chip by 2030. (10 qubits is 210 or 1,024 simultaneous calculations, and 30 qubits or 230 is 1 billion 73 million simultaneous calculations). All this is leading one day to a global-scale quantum Internet. Quantum computers will be able to train and run complex AI models, while AI algorithms can make quantum computing more efficient.
As the Fourth Industrial Revolution evolves, all elements of a business will become connected with artificial narrow intelligence; companies will increasingly become collective intelligence systems. Financial services and other kinds of businesses could just become software. Some unemployment impacts of narrow AI are being seen today, but if artificial general intelligence can be created, then the big impacts on unemployment, economics, and culture will be much greater. Facebook (now Meta) closed down AI bots that created their own language that humans could not understand, and Google’s AutoML can create new AI better and faster than humans, using layers of neural networks. Europol has created NoMoreRansom to decrypt and unlock computers under ransom attacks. Digital twins (advanced simulations) are being created for aircraft, cities, ecosystems, and eventually humans and AGI / IoT to test actions, evaluate impacts, and make improvement with the ability to change the twin through time with new data.
Who owns the intellectual property produced by AI with the participation of many inputs from humans and sensors around the world? How can standards, certification, and testing keep up with AI when humans will no longer know completely how it works? Meanwhile, the Covid pandemic has accelerated the growth of tele-everything worldwide from tele-education and tele-conferences to tele-work and tele-medicine. There are 1.35 mobile billion accounts with transfers of $1 trillion per year. Global telemedicine grew from $18.2 billion in 2016 to $87.8 billion in 2022 and is expected to reach approximately $285.7 billion by 2027. Tele-education is forecast to grow from $250 billion in 2020 to $1 trillion by 2027. Massive Open Online Courses (MOOCs) have enrolled 220 million students in 950 universities offering nearly 20 thousand courses over the past decade (data not including China); 40 million new students signed up for at least one MOOC in 2021.
Read more on millennium-project.org
