Visions of the Future

01 — The Tool and the Trap

The Tool and the Trap

We have all seen or used Artificial intelligence [AI] technology in one shape or form, whether it's using ChatGPT to help write an essay, watching artificially created videos on the news or social media, or even using apps and websites that were coded with AI; it has become a staple of daily human life. As game changing as this technology is, it comes with harmful consequences that could cause devastating ecological and cognitive damage across the globe. By leaning too hard on automating tasks, we risk robbing our brains of chances for deep thinking, losing our ability to think critically, ultimately "shrinking" our brains over time. While at the same time, increasing the pressure on our planets environment through every step of developing AI to having infrastructure to support the growing industry. We also face a problem with people using AI to create videos. These video creation models have gotten to a point where by uploading a simple paragraph, someone can get an incredibly detailed and genuine looking picture or video, easing the ability to spread false information. If we learn how to properly incorporate this technology into our everyday lives, using AI to support creativity, deep research, and productivity, we can not only reverse environmental damages but help humanity reach scales considered impossible to achieve.

02 — What is a Type I Civilization?

What is a Type I Civilization?

The Kardashev scale created by Shkala Kardashyova [шкала Кардашёва] in 1964 is a "method of measuring a civilization's level of technological advancement based on the amount of energy it is capable of harnessing and using." ("Kardashev Scale") according to Wikipedia. On this scale, Kardashyova broke civilizations into 3 major categories, a type I civilization is able to access all the energy available on its planet and store it for later use, a type II civilization can directly consume and store a star's energy, likely through the use of a Dyson sphere [A machine designed to collect most of, if not all, of a suns total energy output], and a type III civilization is able to capture all the energy emitted by its galaxy, including every object within it, such as every star, black hole, moon, planet, down to the last meteor. Earth nowadays would fall at around a type 0.7 civilization, because humanity relies on inefficient sources for our planet's energy, the same sources of energy causing record breaking damages that could even be seen here in Boston, where scientists at UMass estimate "20 to 30% increase in precipitation…up to a 200 cm rise in sea level…and projections of 25-55 days per year above 90 degrees Fahrenheit from 8-10…all by the year 2100." (Douglas and Kirshen 2-4) This environmental strain and technological dependence all affect the direction of humanity's future and our relationship with AI could mark the point where we either evolve into a planet capable of sustaining itself or collapse under our own innovation.

03 — The Environmental Costs

The Environmental Costs

The process of creating and keeping an AI involves the consumption of a lot of valuable resources. The EPA found that "In the case of the high-tech industry, which uses considerable quantities of water to manufacture semiconductors and other components, water is vital to industry operations. Cleaning and rinsing silicon chips can require billions of gallons of water per year; to produce a single chip can use up to 7,900 gallons" (United States, Environmental Protection Agency) The EPA's point makes it clear that chip manufacturing is not just an energy problem but a water intensive process that could drain local supplies at a scale most people don't realize. Training models that are stored on those chips, because of the massive demand for energy, emits a significant amount of carbon dioxide emission and usually over half of all that energy comes from burning fossil fuels (Guidi 6). According to Kashish Mistry in The Carbon Cost of AI, "Training one large language model [LLM] emits approximately 300,000 kgs of CO2." That's more than 150,000 kgs of pollutants added to our atmosphere for a singular training session and more than one is needed to have a final product that would be released to the public.

Building data centers to keep those models up and running also adds to the environmental pressure. Up front emissions consist of construction materials, IT equipment, manufacturing vehicles, supply chain, and even crew transportation, it all adds up. In addition to those, there are also continuous operational carbon emissions from powering servers and systems. A study led by Gianluca Guidi from Cornell University found out that "Data centers—warehouses containing thousands or millions of computing core that serve as the backbone of the modern information technology infrastructure…are energy intensive facilities, with computational power and cooling as the most energy-hungry processes…we identified 2,132 data centers located throughout the contiguous US that were operating from September 2023 till August 2024…we found an estimated aggregate electricity consumption of 192.64 Twh for the 2,132 data center during the study period, assuming a constant uptime or a capacity utilization rate of .075 throughout the year." (Guidi 1,2,3,4,5) to put this massive figure into perspective, in one year these data centers consumed the same amount as a developed nation like Spain or Italy, considering the power consumption for all homes in Massachusetts for one year is around 21 Twh, the energy used for data centers could power all those homes for almost an entire decade.

3,000,000 homes × 7000 kWh = 21,000,000,000 kWh  |  192.64 TWh / 21 TWh = 9.17 Years

04 — The Cognitive Cost

The Cognitive Cost

Artificial intelligence does not just impact the environment; it also reshapes how we think. By depending on AI to solve and understand problems, fully draft papers, or make daily decisions, we risk losing one of the most important parts of being a human being, the ability to think deeply and originally. The more we offload mental effort to algorithms, the more passive our minds become. Researchers at the MIT Media lab refer to this as "cognitive offloading", "We assigned participants to three groups: LLM group, Search Engine group, Brain-only group, where each participant used a designated tool (or no tool in the latter) to write an essay. In the 4th session we asked LLM group participants to use no tools (we refer to them as LLM-to-Brain), and the Brain-only group participants were asked to use LLM (Brain-to-LLM)…We used electroencephalography (EEG) to record participants' brain activity in order to assess their cognitive engagement and cognitive load, and to gain a deeper understanding of neural activations during the essay writing task… EEG analysis presented robust evidence that LLM, Search Engine and Brain-only groups had significantly different neural connectivity patterns, reflecting divergent cognitive strategies. Brain connectivity systematically scaled down with the amount of external support: the Brain‑only group exhibited the strongest, widest‑ranging networks, Search Engine group showed intermediate engagement, and LLM assistance elicited the weakest overall coupling. In session 4, LLM-to-Brain participants showed weaker neural connectivity and under-engagement of alpha and beta networks; and the Brain-to-LLM participants demonstrated higher memory recall, and reengagement of widespread occipito-parietal and prefrontal nodes." (Kosmyna et al. 1). When we let AI do all of our creative and analytical work, our brain adapts by doing less of it, over time that could lead to what researchers call "intellectual atrophy", where curiosity and independence slowly fade away; if we keep letting this happen we lose a lot more than just human creativity, we lose the opportunity to decide the direction of our future. Learning to use AI as a collaborator rather than a cheat code is the only way to protect the part of intelligence that still belongs to us entirely.

05 — Planetary Damages

Planetary Damages

The study led by EllEn Douglas and Paul KirshEn at UMass Boston found that "most projections point to a 10 to 20% increase in daily precipitation intensity by 2050 and a 20 to 30% increase by 2100, depending on the climate model and map scaling approach…The Northeast has already exhibited notable increases in extreme precipitation. For example, the region that encompasses New England has experienced a greater than 70% increase in the heaviest 1% of daily precipitation events over the period 1958 to 2010, which represents the highest regional increase in the U.S." (Douglas and Kirshen 2, 19). Although this study already found that extreme precipitation in the Northeast has increased sharply since the 60s, it uses broad data that does not consider the AI boom and increased demand for power hungry data centers, meaning the Northeast could see even heavier, more extreme rainfall events soon. In other words, the physics models behind these projections make it clear that the energy demand from AI infrastructure has the potential to worsen the climate in surrounding areas, exceeding results.

Heat is also becoming a growing climate risk in Massachusetts, and the shift is already visible in the long term temperature records, projections from the UMass Boston report estimate that "The average number of days over 90 °F historically ranges between 8 and 10 days per year, depending on the county…The BRAG reported the full range of projected days over 90 °F as 25 to 90 days per year… The BRAG expected range of days over 90 °F was 30 to 70 days per year as the century progresses" (Douglas and Kirshen 3). This is not just warmer summers, more heat means more energy demand, more strain on infrastructure, more danger for people living in dense, underdeveloped, already overheated neighborhoods.

Sea level rise is the part of climate change that creeps up slowly but hits the hardest once it arrives, especially in said neighborhoods, and for a coastal city like Boston; damages could get into hundreds of billions of dollars. The report shows how wide the range of outcomes is, depending on how fast we cut emissions. "Under the most optimistic scenario, sea level rise in 2100 compared to a 2000 baseline is 35 to 78 cm versus 72 to 146 cm for a more extreme scenario. Under the RCP8.5 study, 2 m of sea level rise in Boston Harbor is possible by 2100, in 2200 the likely range of increased sea level is 184 to 378 cm." (Douglas and Kirshen 4). That is the kind of rise that permanently changes coastlines, flood zones, roads, subways, and utilities, driving up maintenance costs just to keep basic infrastructure running. Once you factor in storm surge on top of the baseline rise, you are looking at a completely different map of the city and many large scale projects to help adapt to the dangerous weather, creating a massive tab that taxpayers would most likely have to burden.

06 — Pivot

Pivot

AI will keep digging us into a deeper hole, limiting potential for future generations instead of expanding it. Yet I do not see AI as the enemy; it is a tool, one powerful enough to accelerate collapse or accelerate progress but with great power comes great responsibility. When guided by ethics, sustainable design, and human creativity, AI could operate as the missing system that restores the balance we have lost and push humanity toward the next step of becoming a true Type I civilization.

Sustainability is all about meeting today's needs without jeopardizing tomorrows; to reach a Type I civilization, we must use our planet's energy efficiently. It is all a balance between progression and preservation like a closed loop, reusing, regenerating, and making smarter decisions with resources we already have. This is where AI shines, instead of humans trying to manage chaotic grids like nuclear reactors, water pipelines, even factories, AI could fully manage systems in real time, reducing waste on a planetary scale.

07 — Road to Type I: Already Doing

Road to Type I: Already Doing

One of the biggest ways AI is already helping is by cutting down emissions in transportation. According to the EPA, "In 2022, direct and indirect greenhouse gas emissions from transportation accounted for 29% of total U.S. greenhouse gas emissions, making it the third largest contributor of U.S. greenhouse gas emissions when considering indirect emissions from distributed electricity; From 1990 to 2022, total transportation emissions from fossil fuel combustion increased by 19%." By implementing the right tools, we could not only decrease the rate of emissions growth but bring total emissions to a net zero; Google is doing exactly that with their Maps service, using AI to suggest more efficient routes that reduce emissions by minimizing distance travel and fuel consumed. It also considers weather conditions, real-time traffic data and even makes routes that avoid steep hills, micro savings that snowball to around one million tons of CO2 saved per year.

Large language models are composed of hundreds of millions, if not billions, lines of code; like neurons in a human brain, they can connect similar things to make new discoveries in the blink of an eye but unlike humans, some algorithms can compress years of research and analysis into months or even days. This technology is perfect for the drug industry, where a single molecule can decide whether to extend a life or end one. "The vast chemical space, comprising greater than 10^60 molecules, fosters the development of a large number of drug molecules. However, the lack of advanced technologies limits the drug development process, making it a time-consuming and expensive task, which can be addressed by using AI. AI can recognize hit and lead compounds and provide a quicker validation of the drug target and optimization of the drug structure design…The process of discovering and developing a drug can take over a decade and costs US$2.8 billion on average. Even then, nine out of ten therapeutic molecules fail Phase II clinical trials and regulatory approval…Several biopharmaceutical companies, such as Bayer, Roche, and Pfizer, have teamed up with IT companies to develop a platform for the discovery of therapies in areas such as immuno-oncology and cardiovascular diseases" (Paul, Debleena et al.) With the ability to speed up workflows that have breached what previously thought was possible, algorithms can speed up delivery of cutting edge medicine by tenfold, evolving humanity into a stronger, less disease prone, healthier people and saving a ton of money on the way.

The United States relies heavily on other countries for several critical minerals, including many rare earth elements, reducing this dependence on imports is essential for securing supply chains in the advanced manufacturing, defense, energy, and medical industries. "In 2018, as the need for rare earth elements (REEs) and other critical minerals increased, the Office of Fossil Energy and Carbon Management (FECM) tasked the National Energy Technology Laboratory (NETL) with finding more domestic sources… The goal was not just to find new sources of critical minerals. The real challenge was figuring out how to locate them quickly, assess the form and type of each deposit, and transfer that knowledge to the people responsible for evaluating sites in the field. That information could then be used to inform decisions about how best to extract and process those resources…Through this collaborative project, the NETL-led team created an advanced method to rapidly analyze geological data. When coupled with their AI-powered prospectivity forecasting model, it enables users to more efficiently and systematically assess potential sites for critical minerals in unconventional sources…To further improve the process, the team developed a standard way to identify and measure critical elements in geological materials, both in the lab and in the field. Combining this method with AI-informed analysis significantly improves the accuracy of resource estimates and is already being used in research and commercial projects." This project is a promising step for the future of mining, saving time, money, and most importantly stopping pollution due to failed dig sites from causing more damage. This technology has the potential to transform how the United States discovers critical minerals, only coming to life because of government funding during research & development. When public investment, scientific talent, and private industry pull in the same direction, extraordinary projects like this can happen. With the right tools and support, innovation doesn't have to take decades.

08 — Can Do

Can Do

If AI and engineering can speed up drug discovery and optimize entire industries, the next logical step is to look beyond Earth for cleaner, smarter ways to get the materials we rely on every day. Asteroid mining is one of those things you hear from a Futurama episode, thought to only be possible in theory, but AI can bridge the gap between imagination and tangible reality. "Asterank, which measures the potential value of over 6,000 asteroids that NASA currently tracks, has determined that mining just the top 10 most cost-effective asteroids–that is, those that are both closest to Earth and greatest in value–would produce a profit of around US$1.5 trillion. There is also great potential for further expansion. One asteroid, 16 Psyche, has been reported to contain US$700 quintillion worth of gold, enough for every person on earth to receive about US$93 billion… Such technology could also have a tangible environmental impact. Most notably, asteroid mining would prevent the need for traditional in-the-ground methods of mining, which release toxic chemicals such as lead and arsenic into waterways and contribute to acid mine drainage. Asteroid mining could also provide an avenue for the creation of solar power satellites, a potentially consistent source of clean energy… Additionally, an important argument can be made that asteroid mining would reduce the prevalence of inhumane or otherwise illegal practices surrounding human mining operations. This would especially impact artisanal and small-scale mining (ASM) operations, operations that are not managed by larger mining companies. For example, recent attention has been focused on the Democratic Republic of the Congo. This country has responded to the growing global demand for batteries and electric vehicles through its cobalt supplies, of which it contains about 70 percent of the world's resources. Although mining operations can be dangerous, a deplorable record of child labor and fatal accidents within Congolese ASM operations has highlighted the need for significant change." (Yarlagadda). Government funding during research & design phase of projects that benefit the population has proven to work very well, shown from the efforts of the Office of Fossil Energy and Carbon Management and the National Energy Technology Laboratory; It's not just about money, it's about building a system where we can get the resources we need without poisoning ecosystems or putting vulnerable communities at risk. This is the kind of off world resource collection that will push us to a more advanced future. Combining the massive amounts of rare earth materials from asteroid mining, we could ramp up production of solar power satellites and large orbital arrays that collect far more energy than anything built on Earth. These systems are the foundation for megastructures design to capture and manage energy on a stellar scale. Shant Baghram from Cornell University wrote an interesting paper about the possibility and physics behind these megastructures, "An AI-based civilization, as a continuation of intelligent life, will be distinguished from its ancestors by the capability of using energy effectively and vastly. We assume that the ET-AI, at least, is in a second-level civilization on the Kardashev scale. ET-AI uses space megastructures such as Dyson spheres for harnessing the energy of their host and beyond. On the other hand, an AI-based society is computational with processing systems, which need a low temperature for high performance and, consequently, an enormous amount of energy. Accordingly, they will make space constructions like Dyson spheres at a distance from the host star or black hole to use all energy of the source". All together, these ideas show how AI, sustainable engineering, and human creativity, can make a path from today's challenges to tomorrow's possibilities. We may be very far from a Dyson sphere, but the logic, math, and humanities potential all point towards a future where massive scale projects can be possible.

09 — Wrap Up

Wrap Up

Writing this paper didn't just show me what AI is capable of; it reminded me why I try to use it carefully in my own life, I came into this knowing every model we train uses water, adds chip manufacturing waste, massive amounts of energy demand from data centers, and how easy it is for people to offload tasks until their critical thinking has evaporated. It also made me look at my own engineering projects in a different way; when working on my AI-powered motorcycle smart helmet, I'm trying to solve the same problem I've been writing about this whole time: how do you use AI to help people without adding more strain to the world? And with my swarm based solar drones, the whole goal is to build clean energy setups that barely touch the environment, drones placing panels, auto leveling mounts, real time weather sensing, everything coordinating together so we can expand power without tearing up the delicate planet. These projects are obviously nowhere near the scale of asteroid mining or Dyson spheres, but the philosophy behind them is the same; Build systems that protect people and lighten the load on the planet instead of adding to it. What I learned is that this balance is everything. AI is powerful enough to accelerate the climate crisis or accelerate the solutions to it, depending entirely on where we step and what path that choice takes us. When AI is used thoughtlessly, it drains resources, increases emissions, and chips away at human thought. Understanding that tension makes me even more certain about the kind of engineer I want to become. We might be a long way from a Type I civilization, but the decisions we make with technology today are the first steps on that path, and I want the projects I create to push in the right direction.