Teaching Machines to Think: A Journey Through AI, ML, and Beyond
Artificial Intelligence
The concept of Artificial Intelligence can date back to centuries but was officially introduced in more modern times as an amalgamation of separate disciplines. Dating back to the summer of 1956 at Dartmouth college [35], this science was borne from that of Computer Science, Psychology, Mathematics Neuroscience and Physics. This is important to note, as the upcoming topics reveal significant crossovers between disciplines. The notion of Artificial Intelligence can really be defined as the science of teaching a machine reasoning and with an understanding of the world. Not necessarily a one-to-one human machine match, but a way to have a machine perceive and learn the real world. It is essential to note that there is a difference between the techniques being applied to AI - like Machine Learning - and the architecture of some of these - like Artificial Neural Networks.
Figure 1. Artificial Intelligence ML Snapshot
Machine Learning
While AI is often equated with all things, smart machines, it is actually the overarching science behind them, not the specific techniques or technologies used––depicted in Figure 1. Machine Learning (ML) is one technique that attempts to indirectly train machines in a way that provides them with reasoning capabilities without the action of implicitly programming the world. For instance, a traditional software engineer could program steps needed to traverse a maze. For instance, programming the actions of two sequential iterations of up, down, left, right, etc., followed by another three going left, right, up, up. The cost to travel each path could be stored in a large 2D-matrix and then additional operations can be performed from there. Upon re-launch of the program, the same outcome would occur with updated the 2D-matrix with the same outcomes. With this approach, the machine doesn’t learn anything knew – realistically anything at all – about the world.
This just another set of actions that the machine would take, making the end-user appear that it may possess some form of intelligence, which it does not – although intelligence is typically based on human capabilities and is inherently bias. In contrast, the goal of ML in this context, aims to instead provide constraints about the world, thus allowing the machine to make its own assumptions of how to traverse a maze with. This can be achieved with different techniques or algorithms, such as Breadth-first, A*, Depth-first, to name a few, with back-propagation that updates the weights as needed. Running these scenarios for a defined number of iterations - called Epochs – allows the machine to start to understand the world, mostly on its own.
Artificial Neural Networks and Deep Learning
The next more popular idea to this approach is building essentially a reasoning brain, called an Artificial Neural Network (ANNs)––often just called Neural Networks––depicted in Figure 2. This could be thought of as designing a set of probabilities that perform a set of operations at the individual level, often called a node or neuron. Although sometimes misrepresented in mainstream media, these were not meant to be a one-for-one mapping of the human brain and how our brain processes, stores, and accesses memory. Neuromorphic Computing is much closer to that attempt, which will be touched upon later in this paper and the subsequent papers. Neural Nets are built in multi-layers, which consist of an input, output, and a middle layer – referred to as a hidden layer. You may have heard of Deep Learning or Deep Neural Nets; essentially, it’s the same, but will have a minimum of three layers - which includes the output layer, not the input. Illustrated in figure 2, gives an idea of how information would flow from the input to the hidden, and ultimately ending at the output layer.
Figure 2. Conceptual Artificial/Deep Neural Network
Natural Language Processing (NLP)
Natural Language Process (NLP) is a concept and subfield of AI, intent on understanding sentiments withing languages and explores how it evolves over time. This balances linguistics with computational data. Moreover, the get an understanding of how sentiment analysis is performed, the following is basic process:
Large text is collected and ingested, referred to as a corpus.
Preprocessing words –
Normalizing Text – typically converting all words to lowercase.
Cleaned - removing all extra spaces, special characters, clearly misspelled words.
Stemming or lemmatization – changing words like running, runs to run.
Removal of Stop-words – removing high volume words that don’t add much value, i.e. “the”, “is”, etc.
Text is tokenized – converting text into numerical values.
Splitting data – typically splitting the dataset into training, validation, and testing.
After the above the steps are performed, this gets run through a model, which could be a Jupyter Notebook.
Large Language Models (LLMs)
Next, we get into Large Language Models or LLMs. This essentially is a subset to NLP and instead of a very well-defined and focused purpose, LLMs are more of a general purpose, with a less-defined structure. LLMs require much more computational resources and require a lot more data. Traditional NLP is best used for smaller research topics, whereas LLMs, are best suited for broader applications and are trained on essentially all publicly accessible and applicable data from the internet. This an oversimplification, but you can think of LLMs as a broad bigger brother of NLP.
Large Quantitative Models (LQMs)
Expanding on the idea of NLP and LLMs, we have Large Quantitative Models or LQMs. Unlike LLMs, LQMs are composed of more scientific data and with an understanding for the environment. These are typically built by interdisciplinary teams consisting of scientists, engineers, and data experts. This approach well defines the structure using accurate data, equations, and historical data. As an example, you can define the model with something like genome sequencing. Data of existing structures can exist, with examples of what types of biosignatures to look for and can analyze numerous permutations of these protein structures. Unlike LLMs, LQMs are going to be the way of the future within the scientific communities, taking the best parts of the NLP/LLM concepts, and paring that with the bright scientific minds.
References
Lakdawalla, E. (2018). The Design and Engineering of Curiosity: How the Mars Rover Performs its Job. Springer.
Manning, R., & Simon, W. (2014). Mars rover Curiosity: An Inside Account from Curiosity's Chief Engineer. Smithsonian Institution.
Ehrenfried, M. (2022). Perseverance and the Mars 2020 mission: Follow the Science to Jezero Crater. Springer Nature.
Crumpler, L. (2021). Missions to Mars: A New Era of Rover and Spacecraft Discovery on the Red Planet. HarperCollins.
Lee, K. (2018). AI superpowers: China, Silicon Valley, and the New World Order. Houghton Mifflin.
Tyson, N. D. (2022). Starry messenger: Cosmic Perspectives on Civilization. Henry Holt and Company.
Tyson, N. D. (2017). Astrophysics For People in a Hurry. W. W. Norton & Company.
NASA. (2022). Mars Pathfinder mission overview. Retrieved from https://mars.nasa.gov/MPF/rover/mission.html
NASA. (2022). Mars exploration historical log. Retrieved from https://mars.nasa.gov/mars-exploration/missions/historical-log/
Patel, D. (2023). Artificial intelligence & generative AI for beginners.
11. Kelleher, J. (2019). Deep learning. MIT Press.
12. Committee on Scientific Use of the Radio Spectrum. (2010). Spectrum management for science in the 21st century. National Academies Press.
13. Kissinger, H., Schmidt, E., & Huttenlocher, D. (2022). The Age of AI and Our Human Future. Little, Brown and Company.
14. Sheehan, W., et al. (2021). Discovering Mars: A history of observation and exploration of the red planet. University of Arizona Press.
15. Zhang, S., et al. (2020). Deep SCNN-based real-time object detection for self-driving vehicles using lidar temporal data. IEEE Access, 8, 76903–76912. https://doi.org/10.1109/ACCESS.2020.2984855
Quigley, C. (2023). Simply Artificial Intelligence. DK.
Morden, S. (2022). The red planet: A Natural History of Mars. Pegasus Books.
Domingos, P. (2018). The master algorithm: How the quest for the ultimate machine will remake our world. Basic Books.
Kaku, M. (2023). Quantum supremacy. Doubleday.
Kaku, M. (2023). The God equation: The quest for a theory of everything. Double Day.
Tyson, N. D. (2023). To infinity and beyond: A journey of cosmic discovery. National Geographic.
Madi, M., et al. (2024). Artificial intelligence for space: Trends, applications, and perspectives. CRC Press.
Leracitano, C., et al. (2024). The use of artificial intelligence for space applications: Workshop at the 2022 International Conference on Applied Intelligence and Informatics. Springer.
Heuer, R. (2018). Psychology intelligence analysis. Military Bookshop.
Schmidt, E., et al. (2021). Final report | National Security Commission on Artificial Intelligence. Retrieved from https://www.nscai.gov/
Ferrie, C. (2023). Quantum Bullsh*t: How to Run Your Life with Advice from Quantum Physics. Sourcebooks.
Zubrin, R. (2024). The new world on Mars: What we can create on the red planet. Penguin Books Limited.
Chodos, A., et al. (2023). Ghost particle: In search of The Elusive and Mysterious Neutrino. MIT Press.
Weinersmith, K., & Weinersmith, Z. (2023). A city on Mars: Can we settle space, should we settle space, and have we really thought this through? Penguin Press.
Jackson, T. (2020). The periodic table: A visual guide to the elements. DK Publishing.
Cox, B., & Forshaw, J. (2022). Black holes: The key to understanding the universe. Mariner Books.
Christian, B., & Griffiths, T. (2016). Algorithms to live by: The computer science of human decisions. Henry Holt and Company.
Loeb, A. (2023). Interstellar: The search for extraterrestrial life and our future in the stars. Houghton Mifflin Harcourt.
DK. (2020). How the brain works: The facts visually explained.
Webb, A., & Hessel, A. (2022). The genesis machine: Our quest to rewrite life in the age of synthetic biology. Blackstone Pub.
Crawford, K. (2021). Atlas of AI. Yale University Press.
Susskind, L., & Cabannes, A. (2023). General relativity: The theoretical minimum. Basic Books.
Carroll, S. (2024). The biggest ideas in the universe: Quanta and fields. Dutton.
Tyson, N. (2024). Merlin’s tour of the universe: A traveler’s guide to blue moons and black holes, Mars, stars, and everything far. Blackstone Publishing
Chiang, H.-S., Sangaiah, A. K., Chen, M.-Y., & Liu, J.-Y. (2020). A novel artificial bee colony optimization algorithm with SVM for bio-inspired software-defined networking. International Journal of Parallel Programming, 48(2), 310–328. https://doi.org/10.1007/s10766-018-0594-6
Azam, M. A., et al. (2021). Hybrid ant swarm-based data clustering. IEEE World AI IoT Congress (AIIoT), 2021, 0170–0173. https://doi.org/10.1109/AIIoT52608.2021.9454238
Kavyashree, N., et al. (2024). AI in space. Grenze International Journal of Engineering & Technology (GIJET), 10(2), 6297–6302.
NASA. (2022). Communications with Earth | Mission – NASA Mars Exploration. Retrieved from https://mars.nasa.gov/msl/mission/communications/
Barnell, M., et al. (2023). Advanced ultra-low-power deep learning applications with neuromorphic computing. IEEE High Performance Extreme Computing Conference (HPEC), 2023, 1–4. https://doi.org/10.1109/HPEC58863.2023.10363561
Gautam, A., & Sharma, S. (2024). Artificial narrow intelligence inspired neuromorphic computing for logic operations in healthcare appliances. IEEE ICCPCT, 2024, 714–719. https://doi.org/10.1109/ICCPCT61902.2024.10672744
NASA. (2025, January 25). Discover Mars. Retrieved from https://science.nasa.gov/mission/mars-sample-return/
Ni, J., Zhang, Z., Su, B., Fan, X., & Liang, W. (2017). A bio-inspired neural network-based PSO method for robot path planning. 13th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD), 2017, 234–238. https://doi.org/10.1109/FSKD.2017.8393137
NASA. (2025, January 25). Why do we go? Retrieved from https://science.nasa.gov/mars/
NASA. (2025, January 25). Mars exploration. Retrieved from https://science.nasa.gov/planetary-science/programs/mars-exploration/
Omar, A. A., Farag, M. M., & Alhamad, R. A. (2021). Artificial intelligence: New paradigm in deep space exploration. 14th International Conference on Developments in eSystems Engineering (DeSE), 2021, 438–442. https://doi.org/10.1109/DeSE54285.2021.9719425
Ylaya, V. J., Alvarez, A. R., Benasing, P., Gerasta, O. J. L., Caberos, A. B., & Pandian, N. M. (2019). Artificial intelligent (AI) signal processing approach of subsurface electromagnetic radar. 19th International Symposium on Communications and Information Technologies (ISCIT), 2019, 452–457. https://doi.org/10.1109/ISCIT.2019.8905214
NASA. (2025, January 25). Mars Perseverance Sol 1396: Front right hazard avoidance camera (Hazcam). Retrieved from https://mars.nasa.gov/mars2020/multimedia/raw-images/
NASA. (2025, January 25). Mars Ingenuity Helicopter. Retrieved from https://science.nasa.gov/mission/mars-2020-perseverance/
NASA. (2025, January 25). Landing site: Jezero Crater. Retrieved from https://science.nasa.gov/mission/mars-2020-perseverance
Wikipedia. (2025, January 25). Electromagnetic spectrum. Retrieved from https://en.wikipedia.org/wiki/Electromagnetic_spectrum
Intel. (2025, January 25). Taking neuromorphic computing to the next level with Loihi 2. Retrieved from https://www.intel.com/content/www/us/en/research/neuromorphic-computing-loihi-2-technology-brief.html
NASA. (2025, January 25). Closer and closer. Retrieved from https://science.nasa.gov/solar-system/skywatching/night-sky-network/jan2025-night-sky-notes/
NASA. (2025, January 25). Deep Space Network. Retrieved from https://www.nasa.gov/communicating-with-missions/dsn/
Mittal, M., Dewan, P., Panesar, T., & Singh, M. (2024). Educating a robot to walk on Mars using artificial intelligence. 11th International Conference on Computing for Sustainable Global Development (INDIACom), 2024, 1137–1143. https://doi.org/10.23919/INDIACom61295.2024.10498720
Aru, J., Verreycken, E., Laurijssen, D., & Steckel, J. (2022). 3D sonar on Mars. IEEE Sensors, 2022, 1–4. https://doi.org/10.1109/SENSORS52175.2022.9967191
Haj Youssef, K. A., Zayegh, Y., & Alkhedher, M. (2024). Autonomous AI-controlled Mars rover robot. Advances in Science and Engineering Technology International Conferences (ASET), 2024, 1–7. https://doi.org/10.1109/ASET60340.2024.10708660
NASA. (2025, February 18). Mars fact sheet. Retrieved from https://nssdc.gsfc.nasa.gov/planetary/factsheet/marsfact.html
Wilhelm, T., Komann, D., & Wohler, C. (2022). Machine learning on Mars: Open challenges, similarities and differences to Earth remote sensing. IGARSS 2022 - IEEE International Geoscience and Remote Sensing Symposium, 2022, 5373–5376. https://doi.org/10.1109/IGARSS46834.2022.9884096
NASA. (2025, February 18). Taking neuromorphic computing to the next level with Loihi 2. [Photo]. Retrieved from https://www.intel.com/content/www/us/en/research/neuromorphic-computing-loihi-2-technology-brief.html
NASA. (2025, February 18). Deep Space Network. [Photo]. Retrieved from https://www.nasa.gov/communicating-with-missions/dsn/
Anathaswamy, A. (2024). Why Machines Learn: The Elegant Math Behind Modern AI. Penguin Random House LLC.
Benjamin, A. (2015). The Magic of Math: Solving for X and Figuring Out Why. Hatchet Book Group.
Kaku, M. (2014). The Future of The Mind: The Scientific Quest to Understand, Enhance, and Empower the Mind. Anchor Books.
Greene, B. (2011). The Hidden Reality: Parallel Universes and the Deep Laws of the Cosmos. Vintage Books.
Markel, H. (2021). The Secret of Life: Rosalind Franklin, James Watson, Francis Crick, and the Discovery of DNA's Double Helix. W.W. Norton & Company.
Amthor, F. (2023). Neuroscience for Dummies (3rd ed.). John Wiley & Sons, INC.
