Jim Minifie

This from @TuhinChakr is brilliant. That prize winning story from Granta? Turns out it's just a bunch of random whole phrases taken directly from existing text on the internet. Tool allows you to trace those n-grams directly to their source, which is mostly random fanfiction. tuhinchakrabarty.substack.com/p/ai-slop-gran…

🚨 [AI RESEARCH] "Fully Autonomous AI Agents Should Not be Developed" by @mmitchell_ai, @evijit, @SashaMTL & @GiadaPistilli is a MUST-READ for everyone in AI [Bookmark & download it below]: Here's what the authors conclude: "The history of nuclear close calls provides a sobering lesson about the risks of ceding human control to autonomous systems. For example, in 1980, computer systems falsely indicated over 2,000 Soviet missiles were heading toward North America. The error triggered emergency procedures: bomber crews rushed to their stations and command posts prepared for war. Only human cross-verification between different warning systems revealed the false alarm. Similar incidents can be found throughout history. Such historical precedents are clearly linked to our findings of foreseeable benefits and risks. We find no clear benefit of fully autonomous AI agents, but many foreseeable harms from ceding full human control. Looking forward, this suggests several critical directions: 1. Adoption of agent levels: Widespread adoption of clear distinctions between levels of agent autonomy. This would help developers and users better understand system capabilities and associated risks. 2. Human control mechanisms: Developing robust frameworks, both technical and policy level (Cihon, 2024) that maintain meaningful human oversight while preserving beneficial semi-autonomous functionality. This includes creating reliable override systems and establishing clear boundaries for agent operation. 3. Safety verification: Creating new methods to verify that AI agents remain within intended operating parameters and cannot override human-specified constraints. The development of AI agents is a critical inflection point in artificial intelligence. As history demonstrates, even well-engineered autonomous systems can make catastrophic errors from trivial causes. While increased autonomy can offer genuine benefits in specific contexts, human judgment and contextual understanding remain essential, particularly for high-stakes decisions. The ability to access the environments an AI agent is operating in is essential, providing humans with the ability to say 'no' when a system’s autonomy drives it well away from human values and goals." - As AI agents become the latest hyped topic in AI, it's extremely important to understand the risks involved, especially if they are 'fully autonomous' (which seems to be the main goal). 👉 Bookmark & download the paper below. 👉 NEVER MISS my AI governance updates (including must-read papers like this one): join 52,300+ readers who subscribe to my weekly newsletter (link below).

Congratulations to Emeritus @ProfRoyGreen, former dean of the UTS Business School, who has been made a Member of the Order of Australia (AM) for significant service to business, and to tertiary education in the fields of science, technology and innovation.