🤝 AI steals the Nobel Prize spotlight
Physics and Chemistry prizes honor AI pioneers
Dear curious mind,
Welcome to this issue, where I introduce a slight change to the newsletter format. In the AI Update section, you'll now find concise headlines for the latest AI developments, accompanied by my brief perspective on each. For more detailed information on these updates, I encourage you to follow the provided links.
In this issue:
💡 Shared Insight
AI's Nobel Prize: The Dawn of Machine-Driven Discovery
📰 AI Update
Supercharge Your Readwise Notes with the NotebookLM Export
Snipd Enhances AI-Powered Podcast Player with Improved UI and Transcripts
Meta Unveils Movie Gen: Advanced AI Models for Video Generation and Editing
Google's AI Art Revolution Imagen 3 is Now Available For All Gemini Users
🌟 Media Recommendation
Podcast: The Upcoming Industrial Revolution
💡 Shared Insight
AI's Nobel Prize: The Dawn of Machine-Driven Discovery
The 2024 Nobel Prizes have marked a historic moment for AI, with AI researchers claiming top honors in both Physics and Chemistry. This recognition underscores the transformative impact of AI on scientific discovery and raises intriguing questions about the future of innovation.
Groundwork in Physics: Neural Networks Come of Age
The Nobel Prize in Physics 2024 recognizes the foundational work that enabled modern machine learning. John J. Hopfield's and Geoffrey E. Hinton's contributions date back to the 1980s, demonstrating the long-term impact of basic research in AI:
John J. Hopfield invented a network capable of storing and recreating patterns, inspired by physics concepts like atomic spin.
Geoffrey E. Hinton built upon this work to create the Boltzmann machine, which can learn to recognize characteristic elements in data.
These innovations laid the groundwork for today's deep learning algorithms, showcasing how insights from physics can drive advancements in AI.
Applied AI in Chemistry: AlphaFold2's Biological Breakthrough
The Nobel Prize in Chemistry 2024 highlights AI's power to solve longstanding scientific challenges. Demis Hassabis and John Jumper were co-awarded for the creation of the AlphaFold2 system. AlphaFold2 can accurately predict the 3D structure of proteins from their amino acid sequences, a task that previously took years of laboratory work. The system's predictions are freely available, democratizing access to this critical information for researchers worldwide.
Could AI Itself Win a Nobel Prize?
As AI systems become increasingly autonomous and capable of making scientific breakthroughs, an intriguing question arises:
Could an AI one day be awarded a Nobel Prize on its own?
This raises several thought-provoking points:
Attribution of discovery: How do we assign credit when an AI makes a significant finding? Is it the AI itself, its creators, or the researchers who used it?
Nature of creativity: Do current or upcoming AI systems demonstrate the type of original thinking traditionally associated with Nobel-worthy discoveries?
Ethical considerations: What implications would a Nobel Prize for an AI have for human scientists and the competitive structures in research?
While an AI Nobel Prize winner may still sound like science fiction, the rapid progress in the field suggests it's a possibility worth considering. As AI continues to push the boundaries of scientific discovery, we may need to rethink our definitions of intelligence, creativity, and even the nature of scientific inquiry itself.
The 2024 Nobel Prizes serve as a clear indicator that AI has moved from a tool of convenience to a driving force in scientific breakthroughs. As we celebrate these achievements, we must also prepare for a future where the line between human and artificial contributions to knowledge becomes increasingly blurred.
📰 AI Update
Supercharge Your Readwise Notes with the NotebookLM Export (Readwise documentation)
In its core, the new NotebookLM export is just a Google Docs export. Nevertheless, it transforms static highlights into an interactive knowledge base. By enabling you to chat with your curated insights from one or multiple books and articles, it opens up new dimensions of understanding and discovery. The ability to generate the currently hyped AI-powered "Deep Dive" episodes is a nice bonus.
Snipd Enhances AI-Powered Podcast Player with Improved UI and Transcripts (@snipd_app on 𝕏)
The best AI-powered podcast players just got better. Snipd has rolled out a big update which enhanced transcripts and improved the user interface. My highlight is the improved guest extraction, which allows users to easily find and listen to more podcasts featuring the same interviewed guest.
Meta Unveils Movie Gen: Advanced AI Models for Video Generation and Editing (meta.com)
While everyone waits for OpenAI's Sora, Meta's Movie Gen is a welcome competition. The shown results, especially the editing capabilities, look stunning, and the small model sizes would allow running these locally, but the model weights are not released or announced that this will happen. However, Meta shared an extensive technical report with 92 pages.
Google's AI Art Revolution Imagen 3 is Now Available For All Gemini Users (@GeminiApp on 𝕏)
Imagen 3 was already introduced by Google in May 2024, but not accessible for everyone. It produces impressive, high-quality images, though it currently has two notable limitations: it generates only square-format images and does not create images with humans in the free Gemini version. The latter restriction is a safe play against a repetition of controversies Google faced with an earlier release.
🌟 Media Recommendation
Podcast: The Upcoming Industrial Revolution
In a recent episode of the Smart Friends podcast, J. Storrs Hall, renowned AI scientist and author, discusses his vision for the next industrial revolution and the future of technology.
Storrs emphasizes the synergy between abundant energy (likely nuclear), AI, and nanotechnology as the driving forces behind the next major technological leap. He envisions a future with functionally limitless energy, similar to how we view internet bandwidth today.
The podcast explores the "Henry Adams curve" which illustrates the historical trend of increasing energy use and technological advancement. Storrs notes that progress has slowed in recent decades, with energy consumption for each person flattening or slightly decreasing in developed countries.
Storrs argues that energy usage is a better measure of quality of life than economic metrics, stating that more energy consumption generally correlates with a better, safer, and more comfortable life.
An interesting insight shared is that humans tend to travel about an hour a day, regardless of transportation technology. This principle applies whether using walking, cars, or potentially flying cars in the future. The latter is, by the way, a reference to his book from: Where Is My Flying Car?
My take: According to various experts, including the CEO of OpenAI Sam Altman, energy will be the limiting factor for scaling AI further. Abundant energy has the potential to dramatically advance AI and unlock revolutionary technological advancements. The concept of energy abundance could dramatically reshape geopolitics. Countries that lead in developing and controlling these new energy sources may gain significant advantages, potentially altering the global balance of power.
Disclaimer: This newsletter is written with the aid of AI. I use AI as an assistant to generate and optimize the text. However, the amount of AI used varies depending on the topic and the content. I always curate and edit the text myself to ensure quality and accuracy. The opinions and views expressed in this newsletter are my own and do not necessarily reflect those of the sources or the AI models.