John Grant

@jlg

119 Following

162 Followers

John Grant pfp

In human systems, successful protocols are those that navigate the rebound effects their own efficiencies create, rather than those that achieve the greatest efficiency in isolation. I've been working on a protocol lifecycle: 1. Initial protocol design solves coordination problems. 2. Efficiency improvements expand adoption and utility. 3. Partial rebounds occur as usage increases. 4. Complete rebounds emerge in certain domains as protocol efficiencies drive their accelerated exploitation. 5. Protocol debt accumulates as backward compatibility entrenches vulnerabilities. 6. New protocols emerge, or existing weak protocols are integrated, to address the emergent challenges and unintended consequences stemming from the lifecycle of preceding protocols.

0 reply

0 recast

0 reaction

John Grant pfp

Even with perfect information and infinite computing power some things are inherently unknowable https://www.quantamagazine.org/next-level-chaos-traces-the-true-limit-of-predictability-20250307/

0 reply

0 recast

0 reaction

John Grant pfp

Vibe Coding -> Vibe-Driven Development (VDD) -> AI-Driven Specification-Oriented Software Development AlexNet in 2012 based on the CNN architecture that used GPUs and CUDA was the breakthrough moment for deep learning. So, from that perspective it's been just over a decade to get to "vibe coding". It's conceivable AI-driven specification-oriented software development will emerge by 2029, if not before.

0 reply

0 recast

0 reaction

John Grant pfp

AI-driven, specification-oriented development enables "vibe coding" - potentially a punctuated equilibrium in software development. "Specification debt" is the new technical debt. How should we prepare?

0 reply

0 recast

0 reaction

John Grant pfp

Needcasting and Vibe Bargaining https://johnlgrant.substack.com/p/needcasting-and-vibe-bargaining

0 reply

0 recast

0 reaction

John Grant pfp

Been thinking about GenAI from the nature vs nurture perspective. GenAI responses aren't just products of their training data (nurture) but also their fundamental architecture (nature). When Gemini/Claude/etc output different responses based on how prompts are phrased, it shows both aspects at work. So, having more data or querying multiple GenAI models will never overcome the inherent limitations of GenAI to perform true deductive, abductive or analogical reasoning. That's a constraint of their nature, not just their nurture. Just as legal systems aim to provide equal access to justice regardless of privilege or status (though often falling short), I suspect diverse access to frontier AI models should be a fundamental principle too. Different communities asking different questions may not just uncover novel applications but potentially reveal specific practical/ethical limitations that might otherwise remain hidden

1 reply

0 recast

1 reaction

John Grant pfp

In the interest of learning in the open, I've written a response to Simon Wardley's essay 'Non deterministic Languages'. I clarify my position on the distinction between training data and symbolic instructions. While acknowledging our differing perspectives, I also look at the evolution of AI communication through 'strange protocols' and the shift to open science. https://www.linkedin.com/pulse/mechanism-matters-strange-protocols-open-science-john-grant-k4kee

0 reply

0 recast

1 reaction

John Grant pfp

>A closed, state-of-the-art AI model may still be the winner in benchmarks. But an open, state-of-the-art AI model accrues soft power and wins the mindshare of talent everywhere over time – a long-term advantage that cannot be underestimated. That advantage is currently and decidedly tilting towards DeepSeek. https://interconnect.substack.com/p/has-american-soft-power-peaked

0 reply

0 recast

0 reaction

John Grant pfp

The Persistence of AI-Mediated Protocols https://www.linkedin.com/posts/boundaryless-pdt-3eo_protocols-activity-7302659093784088577-TI8G

0 reply

0 recast

0 reaction

John Grant pfp

This interesting Boundaryless Dispatch includes a section that builds on the notion of AI-mediated protocols through the lens of ontologies as "enabling constraints". Their analysis focuses on business implications, how AI agents might transform competitive dynamics by reducing integration costs and weakening traditional business moats. https://www.linkedin.com/pulse/ai-agents-new-competitive-landscape-portfolio-strategy-m48ge/

0 reply

0 recast

0 reaction

John Grant pfp

How long before LLMs, particularly those trained on synthetic data, start to label the field of anthro-complexity as 'Nuancology'?

0 reply

0 recast

0 reaction

John Grant pfp

Problematise: Abstract optimism without a roadmap is just wishful thinking.

0 reply

0 recast

0 reaction

John Grant pfp

Interesting read. But the r2 values 0.32~0.44 indicate there's something amiss, right? Might suggest that human-AI interactions aren't following deterministic patterns. They're emergent and context-specific instead https://www.microsoft.com/en-us/research/publication/the-impact-of-generative-ai-on-critical-thinking-self-reported-reductions-in-cognitive-effort-and-confidence-effects-from-a-survey-of-knowledge-workers/

0 reply

0 recast

0 reaction

John Grant pfp

"… the persistent weirdness of our times. One in which we get free AI from a hedge fund and $200 a month AI from a nonprofit …" https://www.youtube.com/live/zdy9K4UWplM?si=HUN6Qb6Fx3UWFzBG&t=1164

0 reply

0 recast

0 reaction

John Grant pfp

Anyone familiar with the Strategy Evaluation Protocol (SEP)? I'm looking for some pointers or examples of how SEP is currently being used to support Open Science.

0 reply

0 recast

1 reaction

John Grant pfp

From a certain perspective, Open WebUI looks like a mashup between a terminal emulator and a web browser. But I suspect there’s more to it. I think this might be a nascent form of what the next generation of operating systems will look and feel like.

0 reply

0 recast

0 reaction

John Grant pfp

“OpenAI's o3 finds a polynomial time factoring algorithm, none of your crypto assets are safe” https://x.com/realgeorgehotz/status/1881142120171626867

1 reply

0 recast

0 reaction

John Grant pfp

I'm finding Jensen Huang's presentations more entertaining than recent Netflix sci-fi output. https://www.youtube.com/watch?v=k82RwXqZHY8

0 reply

0 recast

0 reaction

John Grant pfp

Should Elon Musk succeed in triggering a sovereign debt crisis, he's well positioned to lobby for X to be the default CBDC platform.

0 reply

0 recast

0 reaction

John Grant pfp

Any sufficiently reductionist explanation is indistinguishable from superstition.

0 reply

0 recast

1 reaction