Transcending the Digital Revolution, AI marks an epic turning point in human history that is challenging the outer limits of our technical, economic, scientific and ethical evolution. AI is so powerful that it can parse billions of data points in a few milliseconds, a task that would otherwise require decades for a human operator to perform. By markedly expanding human boundaries, AI has skyrocketed to the forefront of computer science and radically transformed the way we do business, communicate and navigate the world.
As digital machines march forward, programs like Alexa and Siri have made it easier to interact with AI in the workplace and in the home. AI is also responsible for profound changes in healthcare, education, technology, ecommerce, finance, research, consumer products, manufacturing, transportation, security, arts and media. AI eliminates boring tasks as it has no peer in identifying patterns, entering and analyzing data, generating reports and verifying information.
Along with the extraordinary enhancements attained across nearly every sector, AI has also generated privacy concerns. AI will also lead to mass job displacement, and could very well create bias in hiring, lending and policing. If AI is tasked with improving itself, and if futuristic advancements are not properly aligned, AI could begin to view human thought as substandard and this could spiral into dangerous outcomes. Furthermore, some humans may come to trust AI even when it’s wrong.
Here are some surprising, mind-bending facts about AI that show just how weird, powerful, and unpredictable it can be. Did you know that AI can create its own language and that humans can’t understand it? AI can already beat world champions at games and then coach them. AI can also mimic anyone’s writing style, finish jokes or philosophical musings after hearing just a few sentences. AI is gradually developing the ability to recreate images that people are imagining in their heads. AI art can fool certain judges and win human competitions. AI can also help decode the structure of nearly every protein on earth, a task that would take humans decades.
While we sit back and watch, the AI dilemma begs for answers. Where is AI ultimately headed? What are the goals? To serve humanity? To optimize profits? To enhance national security? To surpass human intelligence? To spread self-serving ideologies? To rule? And who ultimately decides AI’s limits? Presently, there is no single global authority, but researchers and academics are working under the direction of governments, as well as for conglomerates like OpenAI, Google, Microsoft, Meta, Apple, and Amazon, who are all developing profit driven algorithms that are largely unregulated in the business world. These elite players also influence the design, the constraints and goals of the AI systems, decide how AI behaves, what data it is being trained on, what it is allowed to say and do, and where it is deployed.
As AI becomes more autonomous and capable of writing its own code, humanity faces the risk of AI establishing its own limits in ways we may not foresee. Worsening the unpredictability, AI often lacks transparency, as powerful models operate like black boxes, where you can see the input and output, but not what is happening inside. This makes it extremely difficult for AI creators to understand how or why an AI decision was made. There are also many ethical considerations, as AI can create bogus text, images, and videos in order to spread deepfakes, propaganda and misinformation.
Referred to as AI inbreeding, a feedback loop is created by the repeated use of similar models, self-generated content, and by adopting content generated by other AI systems. If you continually copy a copy, models begin to echo each other, leading to self-reinforced bias, intellectual stagnation, loss of quality, and eventual model collapse.
Perhaps the greatest threat is the fact that AI is concentrated in the hands of a few powerful tech companies and self-serving superpowers. The current pace of AI development far exceeds the pace of regulation, disregards any known firewall, and leaves the door open to unpredictable escalation. The United States, Russia and China are resistant to binding regulations because each fears being outpaced by their rivals. Each country suspects that rules will not be followed by the others and that the rogue actors will threaten the momentum they need to develop their own critical assets, economic dominance, and strategic advantages.
As the superpowers sacrifice safety and accountability for speed and strategic advantages, an AI arms race with hacking, weapons and surveillance upgrades is imminent. Once leading nations adopt AI-enabled weapons, others will quickly follow just like they did with nuclear power and chemical warfare agents. AI will undoubtedly have a dual use and be repurposed for military objectives setting the stage for misinterpreted AI decisions that could lead to cyberattacks and hostile engagements. Bad actors will extend AI capabilities to terrorist groups as we are already seeing with drones. In the end, will humans be able to maintain absolute control over AI, or will AI eventually cry wolf and move us to the brink of nuclear annihilation?
AI’s slogan could very well be – “I compute, therefore I AM.” With its superhuman cognition, enriched learning tools and accelerated research abilities, AI holds tremendous promise for humanity. But If AI’s content gets treated as indisputable truth we might enter an era where distinguishing human from machine-made knowledge becomes difficult, if not impossible. AI will require effective oversight and the utmost in human restraint to manage and improve over time. Similar to nuclear arms treaties, AI development needs strict international treaties, mandatory transparency and global monitoring bodies to oversee AI weapons. Responsible humans must remain at the forefront of the decision making loop and guardrails must be put into place to maintain human creativity, values and spirituality.
By Marjan, author of 600 Devils and Fasting Firepower. www.marjanbooks.com
