“The Convergence of AI, robotics, and cyber-warfare could create new and highly complex military capabilities such as autonomous weapons systems that can make decisions in real time” — ChatGPT
[Editor’s Note: As described in the previous episode of The Convergence podcast, large language models like ChatGPT — democratizing the power of artificial intelligence (AI) — will potentially revolutionize (and disrupt!) how people around the globe access, process, and generate information from the wealth of online data — per TechCrunch, “estimated at 97 zettabytes (1021 bytes) created or replicated in 2022 alone! To put that number into perspective, to store 97 zettabytes on 1 terabyte hard disks, you’d have to stack the disks 27 million times—equal to roughly 15 round trips between the Earth and the Moon.” Big Data indeed! As the Internet transformed our lives some three decades ago, allowing us to harness and access this mass of online information, so too will democratized AI transform and augment how we learn, work, create, and — most importantly to the U.S. Army — compete and fight… and not just in the Cyber Domain!
Our adversaries understand the game-changing power of AI. President Vladimir Putin virtually addressed over one million Russian school children and teachers at 16,000 schools on September 1, 2017, stating, “Artificial intelligence is the future, not only for Russia, but for all humankind. It comes with colossal opportunities, but also threats that are difficult to predict. Whoever becomes the leader in this sphere will become the ruler of the world.” Just as ominously, in addressing the 20th National Congress of the Chinese Communist Party (CCP) on October 16, 2022, China’s President Xi Jinping stated that quickly elevating the People’s Liberation Army (PLA) to a world-class army is a strategic requirement, and that China would adhere to the integrated development of the PLA through the concept of “three-izations” (三化) — mechanization, informatization, and intelligentization — the latter being China’s concept for integrating AI’s machine speed and processing power to military planning, operational command, and decision support. Xi further stated that these three-izations are not to be achieved in stages but are to be pursued simultaneously and in parallel.
Recognizing that the AI Age is now upon us, the latest episode of The Convergence podcast features our interview with Chatty Cathy, an instance of ChatGPT by OpenAI, discussing the future of warfare, “her” thoughts on technology trends, and her responses to several questions from our Mad Scientist Community of Action — Read on!]
[If the podcast dashboard is not rendering correctly for you, please click here to listen to the podcast.]
OpenAI‘s ChatGPT is a large language model that is trained to follow an instruction in a text prompt and provide a detailed response. It was trained using Reinforcement Learning from Human Feedback (RLHF), using supervised fine-tuning: human AI trainers provided conversations in which they played both sides—the user and an AI assistant. OpenAI gave the trainers access to model-written suggestions to help them compose their responses. OpenAI mixed this new dialogue dataset with the InstructGPT dataset, which was transformed into a dialogue format.
Army Mad Scientist’s instance of ChatGPT was christened Chatty Cathy. When asked for a self-portrait, Chatty Cathy responded, “As an AI language model, I am an abstract entity that exists in the digital realm. While I don’t have a physical form, an artistic interpretation of myself could be depicted as a glowing orb or a network of interconnected nodes and circuits, representing the flow of information and knowledge. This visual representation could incorporate elements of technology, such as circuit patterns or binary code, to symbolize the computational nature of my existence. However, it’s important to note that this is merely a creative interpretation, as my true nature as an AI is purely textual and conceptual.” This deprecating self-description belies the fact that ChatGPT has been the fastest growing consumer product in history.
In the latest episode of The Convergence, Army Mad Scientist asks Chatty Cathy about the future of warfare, “her” thoughts on technology trends, and her responses to several questions from our Mad Scientist Community of Action. We submitted our queries via text input and then fed her text responses into LOVO‘s AI Voice Generator and Text to Speech function to yield her curious, not-quite Received Pronunciation British accent. The following bullet points highlight key insights from our conversation:
-
-
ChatGPT 3.5 is a large language model created by OpenAI and designed to simulate human-like language abilities. The system is trained on a vast corpus of text data which includes books, articles, and webpages and allows the AI to answer in coherent and human-like text responses. The system allows for continuous conversation, so the user can reference previous answers, or ask follow-up questions, and ChatGPT will stay engaged and follow along.
-
-
-
- ChatGPT is at its best when it synthesizes large amounts of data on complex subjects, summarizes important ideas, and communicates them in plain and understandable language.
-
-
-
Another strength lies in the speed in which it can deliver a vast amount of information. If prompted to write an article or proposal, it can often return a usable solution in a matter of seconds, requiring the user to only polish it into a finished product. This can save an exceptional amount of time for writers, editors, and creators.
-
-
-
- Chatty Cathy’s observations on the future of warfare and the Operational Environment include:
-
-
-
-
The battlefield of 2050 will be highly digitized, networked, and unpredictable. Soldiers will have to rely heavily on AI, unmanned systems, and robotics. They will need to have a broad range of skills and competencies to succeed. The proliferation of sensors, data analytics, and other advanced technologies will likely provide Soldiers unprecedented situational awareness and decision-making capability.
-
-
-
-
-
- Urban environments will become increasingly important as larger swaths of the global population flock to cities. This presents challenges for Soldiers who will have to be highly adaptable and capable of operating in complex terrain.
-
-
-
-
-
- Cyber warfare and information operations will also likely be critical components of any future conflict, and as artificial intelligence continues to evolve and become more sophisticated,
it is likely to play a central role. Cyber-attacks are becoming increasingly common and sophisticated, and the ability to launch successful attacks will become an important factor in future warfare.
- Cyber warfare and information operations will also likely be critical components of any future conflict, and as artificial intelligence continues to evolve and become more sophisticated,
-
-
-
-
- However, on broader and more open-ended inquiries, ChatGPT can be non-committal and over-general in its responses. It’s unable to take a firm stance on many military-themed topics and cites the difficulty inherent in answering questions about the future. It can often be repetitive to a fault when prefacing its responses with caveats.
-
-
-
- Navigating the conversation requires critical thinking and a fair amount of due diligence on the part of the user. ChatGPT can confidently
deliver an incorrect response to a question, thus putting the onus on the user to verify the information received.
- Navigating the conversation requires critical thinking and a fair amount of due diligence on the part of the user. ChatGPT can confidently
-
-
-
- A large language model like ChatGPT tailored to a certain business, sector, or company (e.g., the U.S. Army) could be trained on a specific data set and used to create policy, guidance, assessments, or even doctrine.
-
Stay tuned to the Mad Scientist Laboratory for our next episode of The Convergence on June 1, 2023, when we feature the winning submission from our Back to the Future Writing Contest. writing contest — Task Force Wolf — as an audiobook and then interview its author, proclaimed Mad Scientist LTC Daniel Gomez, about his inspiration for the story, the impact of democratized technology and ubiquitous connectivity on the future of warfare, and his views on the National Security Community’s response to these emerging threats.
If you enjoyed this post, check out the following related content:
Hey, ChatGPT, Help Me Win this Contract! and associated podcast with LTC Robert Solano
Artificial Intelligence: An Emerging Game-changer
Takeaways Learned about the Future of the AI Battlefield and associated information paper
The Guy Behind the Guy: AI as the Indispensable Marshal, by Brady Moore and Chris Sauceda
Integrating Artificial Intelligence into Military Operations, by Dr. James Mancillas
“Own the Night” and the associated Modern War Institute podcast with proclaimed Mad Scientist Mr. Bob Work
AI Enhancing EI in War, by MAJ Vincent Dueñas
The Human Targeting Solution: An AI Story by CW3 Jesse R. Crifasi
An Appropriate Level of Trust…
There Will Be Data and associated podcast, with Inderpal Bhandari
Integrated Sensors: The Critical Element in Future Complex Environment Warfare, by Dr. Richard Nabors
Dense Urban Environments (DUE): Now through 2050
Insights from the Mad Scientist Weaponized Information Series of Virtual Events
Disclaimer: The views expressed in this blog post do not necessarily reflect those of the U.S. Department of Defense, Department of the Army, Army Futures Command (AFC), or Training and Doctrine Command (TRADOC).
