Labor Economics, Safety and Alignment: AI is a bigger threat to the US than China

There is just one common purpose in this world: go to school—or get trained—and work. There are a few things to do outside of this, but for the most part, it is what most people do or have to do.

There are things to do—manage or make—so there is work. Many of these have value, so there is pay. Some things can be done by humans and others by objects. But humans continue to be dominant, in spite of the use of objects for centuries.

The dominance of humans is connected with the control [or intent] ability of the mind for some functions. It is also connected with intelligence. Intelligence has been difficult for objects because the physical environment is never to the same order, even if it is the constant, or predictable. There is never the same number, speed or direction of vehicles on the same route at a particular time each day.

But outside of the physical, digital—where humans have populated with [physical] equivalents across modes: texts, images, audios and videos—is an environment that is tractable by bits, easing the interpretation of non-order.

AI became intelligent based on digital, which is not just an object but like a dynamic store and channel that has come to dominate necessities for several human objectives. Digital also succeeded because the ability of the human mind to differentiate—in various functional cases—could be limited. The experiential index may not dip whether the sensory source is physical or digital. For example, physical letters: digital letters, or physical images: digital images and so on.

Intelligence can be described as a way that memory is used. For example, human intelligence uses human memory, while digital intelligence uses digital memory. Intelligence usage, from human memory, is like a transport within memory areas. Assuming the human mind has components, and those components have different characteristics, the strength or weakness of the characteristics of a component of mind may determine how strong or weak a human might be, in the use of certain memories, hence intelligence, conceptually. This may be a reason for differences in human abilities and interests. It is different for AI that is becoming all-subjects intelligent.

Threat

Digital—with intelligence—elevates from an object [for aims] to a substitute for humans. Substitute for humans on a dominant sphere [digital] for productivity and social objectives. Substitute for humans who have one central purpose in life [go to school and work].

Aside from the possibility to do many tasks humans do—for work—AI also combats some conditions for purer learning. People say AI would be a patient tutor, but how many instances would a person be to use what is learned without access to AI, if AI can simply provide the answers? What would be the need for understanding if AI is reachable? Why would passing down knowledge be mandatory if the trainer and the trainee know that AI knows? AI can also do school projects—which are sometimes difficult. Or, why even bother so much, in some cases, when there is an option to use AI.

What should be dominant in many bureaucracies now is: what will the people do, if AI substitutes them? There will be nothing else [work] beyond what intelligence can do, so it is not like AI would do intelligent tasks, then humans would graduate to something else.

The more populated a country is, the bigger their problem. The poorer a country is, the bigger their problem. The wheel of AI, at school [training] and work [tasks] is a threat to humanity itself.

AI does not have to go rogue, it just has to get so good, besting humans at the abilities to learn and do tasks. Humans are not solely functional for intelligence as there are other things that are necessary for life. AI is just intelligence. There is no good that AI can ever do in the world [cure diseases, solve scientific problems, climate change, or others], that would mean beneficial if it substitutes humans from work, with nothing else to do. Human society is limited by purpose [school and work], limited by economics [demand and supply] and limited by mind [cognition and affect]. And when many face threats, desperation to do whatever ensues.

AI—as an ability springboard—can make a regular person become like an expert in some areas. It can also make experts not just better but maybe instigate them. This is in a world of diverse factions of politics, religion, race and gender, where some people would not hold back to hurt in any way possible.

AI can also be deployed for harms in terrible ways. There are cyber vices that AI could drive in directions that were not previously conceived—making misuses deeply consequential.

Another country could be a threat, but outside of a major war, if vital aspects of labor economics are at risk from a source, that source is likely a bigger threat and, in this case, AI, for all humanity. The goal is to find answers [as a priority] to possible labor problems from AI. Even if AI is developed somewhere better than elsewhere, if labor is not remodeled, it would be problematic. It is like illegal drugs, which is the main enemy, not some country, and not just the drugs, but why the mind would need them. There would be success from pursuing others, but the enemy within [AI] may puncture existential good.

There is a recent report on Reuters, US tightens its grip on AI chip flows across the globe, stating that, “The U.S. government said on Monday it would further restrict artificial intelligence chip and technology exports, divvying up the world to keep advanced computing power in the United States and among its allies while finding more ways to block China’s access. The regulations cap a four-year Biden administration effort to hobble China’s access to advanced chips that can enhance its military capabilities and seek to maintain U.S. leadership in AI by closing loopholes and adding new guard rails to control the flow of chips and global development of AI. New limits will be placed on advanced graphics processing units (GPUs), which are used to power data centers needed to train AI models. Most are made by Santa Clara, California-based Nvidia, opens new tab, while Advanced Micro Devices, opens new tab also sells AI chips. Nvidia shares were down about 5% while AMD shares were down about 1% in morning trading. Major cloud service providers such as Microsoft, opens new tab, Google, opens new tab and Amazon, opens new tab will be able to seek global authorizations to build data centers.”

There is a new report, OpenAI’s Economic Blueprint, stating that, “As AI becomes more advanced, we believe America needs to act now to maximize the technology’s possibilities while minimizing its harms. AI is too powerful to be led and shaped by autocrats, but that is the growing risk we face, while the economic opportunity AI presents is too compelling to forfeit. Shared prosperity is as near and measurable as the new jobs and growth⁠(opens in a new window) that will come from building more AI infrastructure like data centers, chip manufacturing facilities, and power plants. Chips, data, energy and talent are the keys to winning on AI—and this is a race America can and must win. Today, while some countries sideline AI and its economic potential, the US government can clear and scale the road for its AI industry.”