Brain Food

The Circular Nature of AI and Humanity

Throughout history, technological advancements have forced humanity to confront uncomfortable truths about itself. The invention of the nuclear bomb, for example, fundamentally altered global politics. It led to the doctrine of mutually assured destruction, reframing warfare into a precarious balance of deterrence rather than outright conflict. This ushered in the Cold War, an era defined by tension, fear, and the realization that our creations had the potential to annihilate us. AI now presents a similar existential challenge—one that requires us to confront our biases and fears head-on. Read More...

AI Safety Bites

AI Powered Battlefields

Northrop Grumman has recently enhanced its defense systems with new artificial intelligence (AI) technology aimed at improving the speed and efficacy with which soldiers can respond to aerial threats such as drones. This AI upgrade simplifies the decision-making process, allowing for quick responses through a user-friendly tablet interface. The system is designed to handle complex scenarios, facilitating precise and timely actions in response to evolving threats.

This development exemplifies the longstanding relationship between military needs and technological advancements. In my latest article Mirrors of Ourselves, this dynamic is explored, illustrating how the exigencies of war often spur significant technological progress.

You can view Northrop Grumman’s announcement here.

AI Glasses Turned Predatory

The I-XRAY project, developed using META's smart glasses technology, highlights critical privacy concerns by demonstrating how AI can extract personal details such as home addresses and phone numbers just by looking at someone's face. By combining META's hardware with advanced large language models (LLMs) and reverse facial recognition technologies, I-XRAY serves as a powerful example of both the capabilities and risks of modern AI tools in public spaces.

The project offers practical advice on how individuals can protect their privacy by removing their data from online databases. This includes detailed steps to opt out of reverse face search engines like Pimeyes and Facecheck.id, which allow users to request the deletion of their data. Additionally, the project provides links and guidance for opting out of people search engines such as FastPeopleSearch, CheckThem, and Instant Checkmate, which compile public records into easily searchable formats. They also recommend adding two-factor authentication to important accounts and freezing credit as measures to prevent identity theft from SSN data leaks. You can access the project document here.

AI Creative Applications

AI Tongue (not in cheek)

Researchers at Penn State have developed an electronic tongue that uses AI to distinguish between different liquids, such as types of milk, sodas, and coffee, and detect spoilage in juices. This "tongue" employs a neural network to analyze taste through sensors that mimic the human gustatory system. The technology could advance food safety, production, and even medical diagnostics. It offers insights into AI decision-making processes, potentially improving AI applications by understanding the neural network's reasoning.

Read the full article on Penn State's website here.

AI Attribution Progress

Adobe has introduced the Adobe Content Authenticity web app, designed to help creators safeguard their work and ensure proper attribution through Content Credentials. This initiative, part of Adobe’s broader commitment to responsible innovation, allows creators to easily apply secure metadata to their digital work, enhancing transparency and protection against misuse. The web app also provides features to set preferences against the use of their content for training generative AI models, promoting a transparent digital ecosystem. A public beta will be available in early 2025.

Read the full announcement from Adobe here.