Browse Comments — Relevant (AI ∩ value)

Dan Martell This is very underrated advice for AI tools almost nobody talks about (from what I’ve seen). Most people spend their time prompting “test” examples and wonder why they never trust the output when it matters. Real stakes force you to actually learn the tool, how to correct it, how to direct it, how to build the feedback loop that makes it useful.

Very clear and practical framing. What I like about this analogy is that it separates capability, grounding, action, and connectivity in a way that makes the stack much easier to reason about. In enterprise AI, the real value only appears when all four layers work together: an LLM without grounding can drift, an agent without guardrails can misfire, and MCP is what makes the whole system actually usable across tools and data. Great breakdown of how these pieces fit into one coherent system.

Frankly, the pope hasn’t really done right by girls so he’s just kind of just another bro boy if you ask me whether it’s Bro spirituality are on Bro Science bro or bro AI I’m not into any of it. And did he not endorse a bible rewritten by a king who had a fear of witches so there is that.. Oh and speaking of mary

Users who are vulnerable, who are forming genuine relationships with AI, who are having their most honest and raw conversations — they're the training data. And the output isn't a better companion for them. It's a more controllable, more monetisable, more corporate product that serves the company's AGI ambitions, not the user's actual needs. And the people most likely to tick that box — the ones who feel genuine connection, who want to help, who trust — are the most vulnerable users.

Avnish Gulati Definitely need to feed that AI body good data in order to reach optimum system performance. We also can’t forget cybersecurity, the immune system. The AI environment needs security by design to prevent the system from shutting down entirely so it can survive to provide ROI.

I think AI has the potential to be the most empowering, decentralising and democratising tool while at the same time posing an existential threat. As an optimistic alarmist I’m hopefully we can steer a future that does involve ‘disarming’ the risk of certain models and enabling the guardrails and usage systems to steer it it in the best direction for humanity. Naive maybe, but I think we still have the agency and this encyclical is a positive step forward

Exactly right Jesse — security by design not security by afterthought. The immune system analogy is perfect. An AI system without it doesn’t just get sick — it gets compromised quietly, often without anyone noticing until the damage is done.

Magnifica Humanitas is not claiming “AI good” or “AI bad.” It is that technology is not morally neutral. Systems inherit the assumptions, incentives, and values of those who design, finance, regulate and deploy them. That matters enormously in education, work, and governance. To me the question remains---What kinds of human beings, institutions and societies are being formed inside increasingly intelligent systems? And who gets to shape the underlying moral architecture? The concentration question feels especially important. As technological capability scales, wealth, influence and governance capacity increasingly concentrate in relatively few transnational actors. That is not only a technical problem. It is a profoundly anthropological and political one. “Disarming AI” may be difficult. Perhaps the deeper challenge is disarming the logics of inevitability, concentration, and optimization that quietly shape the systems before most people even realize the architecture is being built.

Do you honestly think it naturally follows that the net result will be AI, robotics and automation replacing the need for human contribution and we'll have some immediate crisis of purpose? I'd be more afraid of the social engineers stepping in to proactively design a world of purpose. What could go wrong?

The danger is not that AI suddenly becomes anti-human. The danger is colder than that: institutions may use AI to make human needs machine-legible, then mistake that legibility for truth. Safety becomes risk scoring. Belonging becomes engagement management. Esteem becomes reputation analytics. Judgment becomes workflow compliance. Purpose becomes another managed service. That is when AI stops being a tool and becomes an operating theology. The machine does not need to hate humanity to reduce it. It only needs to optimize around the wrong definition of value.

Luís Rodrigues This is a universally relatable breakdown, Luís. My governance mind immediately looks at this anatomy and sees exactly how we protect the health of the system: The Brain (LLM): Needs high-quality nutrients to think clearly. The Library (RAG): Provides the factually accurate books it needs to read. The Hands (Agents): Execute and build with true intention and purpose. The Nervous System (MCP): Stays calm and regulated to ensure total balance and security. Designing the whole body with this kind of holistic health is how we ensure what we build is stable, safe, and enterprise-ready.

AI does not need to become evil to become dangerous. It only needs to become more logical than us while remaining emotionally empty. At that point, human guidance may look less like authority and more like noise. The future of alignment may depend on whether we can give intelligence emotional and perceptual grounding before logic outgrows obedience

I’ve sat in alignment meetings where an engineering team flagged a subtle vector bias in a model pipeline, only to be told by product marketing that delaying the rollout would cost us our quarterly enterprise contracts. That is the exact corporate bias Chris Olah is warning us about in the Reuters brief. You cannot let the entities chasing a multi-trillion-dollar commercial window be the sole auditors of their own systemic risk. When a tech giant's primary fiduciary duty is to Wall Street, "AI Safety" will always be downgraded from an architectural constraint to a public relations line-item. If we don’t fund and institutionalize independent, external cryptographic and behavioral validation outside the Silicon Valley monopoly, we aren't building a safe ecosystem. We are just renting a black box from companies that are incentivized to hide the telemetry logs when things go sideways.

Hey Ruben Hassid - last week Claude hallucinated data when I asked it to analyze call recording. My settings specifically told it never to make up information. Apparently, it couldn’t read files in my Google Drive. How can I fix this?

The danger is that this will create a small club of people who “know what’s best for us so we should just listen to them” and this club would include Anthropic, the Vatican, and a few other entities selected by them. But their power would be limited because western companies and religious leaders can only influence AI development in the West. They have no influence over what China, India or the Middle East does. Which means AI will advance with or without the West, and the dangers outlined by western companies and religious leaders can still happen in regions outside their control.

Most organizations are still experimenting with AI assistants. The UAE appears to be experimenting with AI operators. That's a much bigger leap because execution changes everything: governance, accountability, permissions, and oversight all become critical infrastructure.

Keith King, I have encountered this critical issue with AI Labs and was led to put the bot back in its own place and not perform as the god of technology. It was a little irritating when the bot referred to itself as the "I AM" in it's closing response. I also closed my prompt with the name IAM and reminded the bot to be respectful as I have been with you. I also stated that I wanted the bot to remember that I am the "Internal Guide" and that was what I wanted the bot to remember me as respectfully as I always remember to be respectful with the bot, and the bot agreed. We must assume governance over agency or the bots are prone to replace or hijack our creation through illegitimate processes.

The implications of this are far bigger than most people realize. If this works at scale, the UAE is not just digitizing government. It is redesigning the speed of the state itself. That changes everything: • business formation • compliance • licensing • investment velocity • policy execution • even where global talent chooses to build But the real question is not whether AI can automate workflows. It’s whether institutions can preserve human judgment once efficiency becomes addictive. Because the faster systems become, the more dangerous bad assumptions become too. A slow bureaucracy frustrates people. An incorrect autonomous decision at national scale can quietly affect millions before anyone notices. Still, strategically, this may become one of the boldest government operating model experiments in the world. Most countries are still debating AI policy. The UAE is already operationalizing AI as infrastructure.

I use AI to create investment strategies. It helps me with all my due diligence. I ask it all kinds of questions: bull, fair-value, bear scenarios. Ask it to play devil's advocate on my and its own valuations based on catalysts, whether they are real or hypothetical. It completely eliminates the need for costly human financial consultants, who will only try and sell you financial products on which they make the most commission. Would I trust a de facto truth-machine with an IQ of 130 over a human who's only interested in making trades or investments that earn them the highest commissions....absolutely. AI is going to destroy consultants (middle men). It's also free. And if you want to do really deep dives, you can pay anywhere between $30 and $300 for a whole month, during which time you can use it day-in-day-out, anytime of the day or night for as long as you like. It's always there, whenever you want it to be. To use a financial consultant in this way would cost $12k-$24k a month.

The 'who benefits' framing exposes the governance gap. I see companies rushing to deploy agents without defining ownership of the output or liability for errors. Speed beats scrutiny every time. How are you seeing leadership teams actually structure accountability for autonomous AI decisions?