JERRY ARCHER
As Artificial Intelligence (AI), the Internet, and technology continue to advance, they raise critical questions about the future of society, culture, security, and privacy. As the media and governments reflect of the impact of latest technology, particularly the public introduction of AI, we must envision the future in a digital world. In this world, the evolution of technology is not only constant, rapid and radical, but has also become a pervasive force driving human evolution.
AI, cloud computing and the Internet make possible the future envisioned by genomics researcher Juan Enriquez in 2012, one where we choose what and who we are. Recent FDA approval of CRISPR (short for “clustered regularly interspaced short palindromic repeats”) technology to treat sickle cell disease in humans is the tip of the iceberg that research scientists use to selectively modify the DNA of living organisms including humans to accelerate research and treat diseases such as cancer and mental illness — we may even decide to use it to change the genomes of our children (i.e. designer babies).
It is technology which allows Intel Fellow Mark Bohr to foresee that “in the future, chips may become integrated directly with the brain, combining AI/human intelligence and dramatically enhancing our cognitive and learning abilities. … lead[ing] to a “technological singularity” — a point in time when machine intelligence is evolving so rapidly that humans are left far, far behind.” Is not the evolution of technology a “butterfly effect,” a change so profound that the world we know today simply disappears?
Cloud computing, or AI, may turn out to very well be the flutter of that butterfly’s wing, altering most everything we know, including ourselves. Noted philosopher Thomas Kuhn cast revolutionary changes as paradigm shifts, where “a paradigm shift [is] a mélange of sociology, enthusiasm and scientific promise, but not a logically determinate procedure, ergo the final outcome of such a change is beyond our ability to discern, and, like the “butterfly effect,” can (and will) have vastly different outcomes whose “first cause” are but tiny differences at inception. Santa Fe Institute professor of economics Brian Arthur says that when complex technology becomes transparent to its users, the complexity vanishes, and it is then that those users assimilate the change. ie: it is then that revolutionary outcomes are possible.
This is happening all around us today in what seems every aspect of our life, just look around! Apple has sold over 2.32 billion iPhones with 97.7 million iPhones shipped in the first two quarters of 2023 alone. It is not too difficult remember 15 years ago when Apple fluttered the proverbial butterfly’s wing with a simple, yet logical paradigm shift, evolving the iPod into the iPhone and thereby creating the smartphone, a profound change in technology and the way in which we touch each other and the Internet — and, to the evolutionary point, how technology touches us.
But, in contrast OpenAI introduced an early demo of ChatGPT with a simple dialogue box that obfuscated the massive complexity of neural networks, large language models and regenerative AI models on November 30, 2022, and the chatbot quickly went viral exceeding 100 million users in January 2023 and more 180 million users as of December 2023, making it the fastest growing technology introduction in history. The company also now has over two million developers, including over 92 percent of Fortune 500 companies working to leverage AI, with more 1.5 billion website visits per month.
Paleobiologist Stephen Jay Gould described evolution itself as “punctuated equilibria” — periods of stasis with little change beyond genetic drift ultimately broken by short periods of radical realignment across many species. The iPhone, CRISPR technology, ChatGPT and other computing capabilities demonstrate the incredible pace of technological change. The question for Gould, were he still alive, would be this: What is the evolutionary implication moving from rare punctuations of long-lived equilibria to constant punctuation of equilibria that are all but evanescent?
Revolutionary evolution is the future — a series of paradigm shifts with unpredictable yet profound effects. Precisely because we are building upon an interconnected foundation of incredibly complex technologies, any small change may be extremely amplified. The Apple App Store, for example, offers 2+ million applications. In dollar terms, that is $1.1 trillion in revenue for Apple and app developers many of whom can be called “cottage industry.” With respect to software development, this is revolutionary evolution — but now there is ChatGPT, et. al., which will likely dwarf the App Store with more than 2 million developers today.
The computing paradigm shift is fueled by near infinite MIPS and storage with near zero marginal cost. Now, when coupled with the Internet of Things, it is clear this will cause fundamental changes. Harvard Business School professor and researcher Clay Christensen’s theory of “disruptive innovation” is just another way of expressing an equilibrium punctuator, a paradigm shifter, but what if that disruptive innovation were to the continent of Africa rather than to an industrial sector? The World Bank and African Development Bank report there are 650 million mobile users in Africa, surpassing the number in the United States or Europe. In some African countries more people have access to a mobile phone than to clean water, a bank account or electricity, the agencies add. All able to touch the Internet not with computers, but with a transducer that obscures the underlying complexity and dramatically alters social interaction on an entire continent and surely the world. The Arab Spring revolutions of 1848, which are sometimes referred to as the “Springtime of Nations”, and the Prague Spring in 1968 could easily pale by comparison.
Philosopher Martin Heidegger, in his 1954 essay “The Question Concerning Technology, describes technology not as something we build, but as something we “uncover and enframe.” The evolutionary change that is upon us is simultaneously centralizing that which has heretofore been decentralized and decentralizing that which has heretofore been centralized, and doing so in the midst of networking everything (the Internet of Things).
Are we simply uncovering and enframing this change? Google has built a quantum computer that’s about 158 million times faster than the world’s fastest supercomputer. D-wave has developed a quantum computer and claims the computing power of 10**29 PCs. IBM demonstrated a technique for increasing the density of storage devices by requiring not 1,000,000 atoms per bit but only 10. If made commercially practical, imagine 100,000,000,000,000 (100 TB) of persistent storage in a single small device. IBM also showed how to make what today would be a supercomputer in the space of a sugar cube. Today’s computing capabilities are only a whiff of what will, according to Kuhn, be unfathomable.
Consider an example of future banking operations. With a personal orbit of thousands of devices an individual would have significant redundancy and thus losing data would be a thing of the past, unheard of in our cloud-based world. It would be in the economic best interest of a bank to store all of your banking data on all your devices, and leverage your computing resource to move, alter and cryptographically ensure the integrity and authenticity of that data. A transaction would be performed by the bank, acting as a trusted third party, moving money from the payor’s device to the payee’s device. The bank would have minimal infrastructure and little IT cost, a significant incentive to move in that direction. Moreover, banks and other financial institutions today already have significant apps that are resident on individual smartphones, so this next logical evolutionary change, while relatively modest, would have profound and revolutionary effects on our financial system, but then there is crypto currency — hmmm! As always, regulators would be so far behind as to be all but out of sight.
On a Doctor Oz television show, the Toto Intelligent Toilet II was characterized as something that would enable you to live longer. The Intelligent Toilet is an Internet connected thing, not your ordinary toilet. Recording and analyzing important medical data like weight, BMI, blood pressure and blood sugar levels, and sending that data to trained physicians, enabled by AI, who can monitor your health and provide early detection for many medical conditions. The segment ended with the line “Trying to have a baby? Not sure when you are most likely to conceive? Ask your toilet for help.” Does that change everything about a trip to the potty? But, that is beginning of the journey. In a generative (an logical advancement of regenerative models) AI model, AI would train itself based on the outcomes of all the billions of connected devices and become better and better at predicting or preventing negative outcomes.
Today, there are over 15 billion devices connected to the Internet, only 20% of which would be characterized as a computer. Such things as smart-phones, televisions, stereos, refrigerators, beds, cars, you name it! Perhaps the more compelling trend is the devices in your orbit are opportunistically communicating to each other. So your fridge can talk to your oven and decide what to make you for dinner based on what is in the fridge, or perhaps in a few months discussing your health with your bed and toilet to make sure you are eating the right things to improve your health, generating a shopping list to your grocery store to make sure you buy (only) the rights things, not that pepperoni pizza which gives you heartburn and causes you to lose sleep, waking up tired and a bit groggy as detected by your smart bedroom slippers or bathroom scale; and inattentive as detected by your car, which gauges your reactions while driving to determine if you are alert enough to drive and if not to signal you in a rude persistent, perhaps interventionist manner to pull off the road and rest or, subtext, stop eating pepperoni pizza before going to bed!
For want of a better term, let’s call this phenomenon “Massive Integrated systems of Smart Transducers” (MIST) the stuff clouds are made of. As we have moved to the cloud we are creating enormous amounts of MIST. Endpoints are more and more not simply computers but transducers with enormous sensing and compute power. Everything from SCADA systems to SIRI on the iPhone, to that intelligent toilet, to Progressive Insurance’s “Snapshot” surveillance device, to 36-row corn planters driven by GPS, to personalized advertising informed by traffic analysis on social networks, to standoff biometrics, to auto-drive cars, to bid-bots on eBay, to …, the complexity of computing is hidden and the user of it willingly assimilated. This change is profound, and now even a small evolutionary change rippling through this complex foundation can cause revolutionary results not only in the realm of technology but also in society, our culture, and our governments and on a global scale.
Not long ago, the California Highway Patrol in an effort to curb speeding acquired reader devices for (the Federally mandated) automobile black box (“Event Data Recorder”), which as it turns out record the car’s speed over some extended period. If you were stopped for any reason the CHiPs would “jack into” your car and then compare your speed with the speed limits and if exceeded would issue you a speeding ticket — a dangerous precedent pointed out in the California Supreme Court ruling prohibiting such actions. But could that ruling be overturned to protect the lives and safety of the driving public. After all (driving is a privilege, not a right), in 2020, 11,654 people were killed in motor vehicle crashes involving alcohol-impaired drivers, accounting for 30% of all traffic-related deaths in the United States. 32 people in the United States are killed every day in crashes involving an alcohol-impaired driver—this is one death every 45 minutes. The annual estimated cost of crash deaths involving alcohol-impaired drivers totaled about $123.3 billion in 2020. Drug-impaired driving is also an important public safety issue, but statistics are not available and clearly add dramatically to the problem. Suppose the CHiPs could interrogate your car remotely, determine if you are driving impaired and tell your car to pull over and disable itself until an officer arrives and with probable cause provided by your car, administer tests to determine the source of impairment and execute an arrest if appropriate. Interestingly, your car could testify against its owner providing an analysis of your condition leading to the probable cause and subsequent arrest. Does the continual increase in the number of deaths change the calculus of the courts? Rental car agencies and maintenance-included leasing deals already use black box data to discipline drivers, reminding us that the word “voluntary” has a meaning in law that bears little resemblance to its meaning in conversation.
Even without directly implanted transducers of the sort described by Bohr, every nuanced movement or change in your orbit can be measured analyzed and correlated. So, while there may be a temporary fear of direct transducer implants, your physical condition, actions and even intentions can be indirectly inferred from mega-sampling large number of interconnected transducers, analyzed by AI models, providing exactly the same result as an implanted transducer.
Dan Geer in a paper remarked “& remember that the Internet was built by academics, researchers, and hackers — meaning that it embodies the liberal cum libertarian cultural interpretation of “American values,” namely that it is open, non-hierarchical, self-organizing, and leaves essentially no opportunities for governance beyond protocol definition. Anywhere the Internet appears it brings those values with it (treating censorship as a routing failure, for example). Other cultures, other governments, know that these are our strengths and that we are dependent upon them, hence as they adopt the Internet, they become dependent on those strengths and thus on our values. A greater challenge to their sovereignty does not exist. The challenge to our sovereignty is its dual — it is the choice of whether to commit our critical infrastructures to the Internet in the entire, to discard our fallbacks along with those who practice them, to bet the farm on a roll of the geopolitical dice.”
While I have tended to a personalized orbit in this essay, there is no doubt that a nuclear power plant is in the MIST and has a myriad of transducers in its orbit, as well as banks, the military, the government, and on and on, many of which will periodically cross paths and exchange data (whether deliberately, opportunistically or maliciously) while in others orbits. In fact, might that be the future of how we learn of new products and services, how our transducers are serviced or upgraded.
This leads me to a few big questions: What is the meaning of security and/or privacy in the evolution of clouds and MIST? How will transducers and apps in your orbit be protected from intrusion, alteration, or service denial? Are laws or regulation needed to protect or govern devices? Would laws or regulation even help or obfuscate the revolutionary evolution happening in spite of sanctioned limitations? Are you in control or entitled to be aware of the apps and transducers in your orbit? Would your apps or transducers have some sort of inferred or legal rights preventing you from turning them off or excluding certain ones? Is security merely a euphemism for control, and if so by whom? Is security evolving into an organic model perhaps one where your transducers have a kind of antibody that roams within your orbit identifying and destroying perceived malicious intruders, and if so what about the impact of false positives? Is everything we understand about our security and privacy about to be obliterated? For that matter, is the meaning of “self” up for grabs? Is security obsolete, collateral refuse of a paradigm shift, replaced by smart “security” bots — like The Matrix?
As Edward Lorenz describes the “butterfly effect,” a slight change in the initial conditions can (will) have vast impact on the outcomes. In the near future are fundamental changes in our perspective and implementation of security and privacy needed? Thought leadership will have a profound influence on the future, enabling a safe and secure paradigm shift into the cloud, AI and MIST. We must not simply let the future happen.
No comments:
Post a Comment