GPT-101 Revisited¶
I spent a good part of my winter break updating the first GPT workshop that I co-designed in 2023. Although the fundamentals of prompt-writing haven’t changed much, the broader AI landscape has exploded.
Each week, I’m staggered by new developments and by the sheer number of AI platforms now available for chat, scholarly research, code generation, and image or video creation. Dozens of competing platforms are vying for monthly subscription fees—often around $20 a month.
Ethics of AI¶
Questions persist about how these AI systems are built, including the critical issue of what data they are trained upon. Meanwhile, there is growing uncertainty about how AI might reshape our daily lives at all levels of society—from the classroom to the courtroom.
Will AI create greater inequality, or can it help lift up disadvantaged groups? Already, we know that:
- AI is now the largest area of financial investment for venture capital.
- Energy demands for AI are skyrocketing, potentially leading to higher consumer costs.
- AI received two Nobel Prizes (in Physics and Chemistry) in October, highlighting the growing influence of computer science in “classical” scientific fields.
Ethical AI¶
Robotics and autonomous systems are entering a new AI-fueled renaissance. Physics-based AI foundation models and synthetic training data are enabling self-driving vehicles and backflipping androids to find their way into both civilian workplaces and military applications. In this context, Isaac Asimov’s Three Laws of Robotics continue to cast a long shadow over the discussion.
While human-piloted drone warfare has already reshaped conflict between Ukraine and Russia, fully autonomous systems inch closer to becoming operational on the battlefield.
AI for All¶
The so-called “Magnificent 7” stocks—the largest companies in the world—are all betting heavily on AI. In the U.S. alone, hundreds of billions of dollars are being poured into new data centers, with grand promises like curing cancer.
While these Fortune 500 companies keep their AI models proprietary, the Chinese startup DeepSeek has open-sourced theirs, trained on commodity hardware. Remarkably, DeepSeek’s technology can run on personal computers or mobile phones, with performance reportedly rivaling models from Google, Meta, and OpenAI.
If these developments continue, it may mean that AI remains accessible even to those without substantial resources. However, as investors increasingly focus on autonomous systems (robotics and androids) and drug development, questions of data extraction and privacy in the U.S. remain largely unregulated. New legal protections, such as an AI Bill of Rights, and broader AI strategies from state and federal governments are sorely needed.