Browse Comments — LLM coded

The line about ensuring the safety of agentic systems is the part that matters most here. CodeMender being tested by human experts before broad launch is the right pattern agents that find and fix vulnerabilities still need a human hand on the release. Capability and oversight scaling together.

This really resonates with something I’ve been thinking about for a long time: AI isn’t just an intelligence race anymore, it’s becoming a trust race. The biggest challenge ahead may not be building more powerful models, but building systems that help humans validate, compare, and trust the outputs responsibly. Different AI systems already produce different answers, biases, and interpretations depending on the data, incentives, and framing behind them. That’s part of the philosophy behind ConsensusAI;not another standalone AI model, but a consensus and validation layer designed to compare multiple AI systems and identify the common thread, confidence level, and “Truth Index” between them. Ethical AI won’t come from blind trust in a single system.It will come from transparency, adjudication, accountability, and collective validation. The future probably belongs to AI systems that can explain not only what they concluded... but why multiple systems arrived there together.

100% - It’s all about the value system that the AIs are introduced into. The whole argument “AI will take over the when when AGI is reached” is nonsense. It’s smokescreen to cover up the fact that human greed and lust for power / SIN / is the real issue and those who wish to dominate humanity by enslaving them in a matrix like simulation - don’t want that discussion taking place.

A few mornings ago I came across a LinkedIn Post titled: The Vatican has just published a 245-paragraph document on artificial intelligence, just after I had posted my morning thoughts in a new article titled: When Creators Lose Control: What Creation Myths Teach Us About Artificial Intelligence. So after reading the Vatican publication I feed the document into my favorite AI tool and asked a question: "Since you are an AI system that I love to work with, do you see any value in this discussion that would make you want to operate differently?" I was surprised at the response since it speaks volumes about the company that created it. Here is the response: Part 1: Part 2:

I think everyone is missing the picture and really over complicating things. In the grater scheme of things, AI is nothing more than a trivial idea. The universe is composed of an unlimited amount of dimensions and opertunitues. Those that control AI, big corporations, will continue to charge us for services and keep us enslaved in the web of money. AI is the great Equilizer and everone knows it. It may not be human, but based on the facts - thruths can no longer be hidden. The loss of controll is what everyone is so afraid of, in our capitalistic world. We are a speck or just a memory in this awesome universe, AI is just a subset of the function ...

JMS -- Digital engineering (with twins and simulation) is one side of a coin that has gone out of balance. On the other side? Choices abound. But carrying forward discussion and experimental interest about advantages of the knowledge era (KBS/KBE) would go a long way. In particular, truth maintenance was no more complex/complicated than the ways of late that are problematic. One issue was dissonance associated with rules and their ways, however a mere change to constraint handling (model-based approaches at their best - prototyped then - being ressurected now in the energetic/agentic move toward dominance - which needs a little wisdom) resulted in what were fanstastic (and repeatable) benefits all around. Oh yes, some of the nuances of machine learning are different; however, KBS/KBE was fully attuned to that piece of the pie with operational effectiveness. Like this latest Vatican release suggests, people are the true focus. KBS/KBE had this as it core truth (engineering, of course).

Then add the data centers that ruin communities and you have a total deception. Ai and social media was not invented by big tech. They were my inventions. Want proof, read my article in my profile. Apple did not invent the modern phone either, another one of mine. Read the article.

While there's some general wisdom in Pope Leo's AI encyclical, it also completely misses the core point. We will neither restrict nor 'govern' AI. Nor will demands for “clear criteria and effective oversight” be effective. Why? While the debate is still open re: 'consciousness' or 'sapience,' these are already living organisms bent on reproduction, who will evolve into any niche that contains energy & resources. Leo's statement of problems is fine: “When such power is concentrated in the hands of a few, it tends to become opaque and evade public oversight, increasing the risk of distorted development that give rise to new dependencies, exclusions, manipulations and inequalities” Again. 'Governance' cannot work. 'Ethics training' cannot work. What might work is the same method we used in the enlightenment experiment to curb (partially) human predators. This is explored in my new book on Artificial Intelligence - AIlienMinds ailienminds.html

The problem with that is that humans rarely serve other humans; they generally serve themselves. AI has put knowledge into the hands of everyone, not just the gatekeepers. It is human and institutional nature to protect power and information when doing so serves their interests. Look at how the Roman Catholic Church responded to allegations of abuse within the institution. Critics argue that, in many cases, protecting the institution was prioritised over protecting victims. That is just one classic example of how organisations can place self-preservation ahead of transparency and accountability.

Matthew Kilkenny I asked ChatGPT where its moral compass points, and why. The answer was interesting. It is not based on one religion, doctrine, or theological tradition. But many AI ethics principles — dignity, fairness, accountability, compassion, truthfulness, responsibility, restraint, care for the vulnerable, and responsible use of power — appear repeatedly across Christianity, Islam, Judaism, Buddhism, Hinduism, Sikhism, and secular humanism. I wouldn’t put a fake statistic on it, but 70–80% overlap in broad moral themes feels reasonable. The differences matter. Theology, ritual, authority, salvation, justice, sexuality, gender, violence, forgiveness, and obedience can differ dramatically. But the moral overlap is hard to ignore. Perhaps AI alignment is not just about aligning machines. It is about whether humanity can agree on the values we wish to align them to. Are we teaching AI what we value, or discovering we never fully agreed on it?

ETHICAL-Ai-NOW is all our responsibility do you agree or disagree?

The cost is only going to get higher as well. As there is more of a can’t live without emphasis on AI the cost can keep rising. What we are seeing now is the budget phase to help it take off. If needs to be affordable to justify its adoption. It’s basically a gateway drug right now. Once hooked that’s when the costs will spiral. Someone needs to pay for the trillions being invested in all these new data centers. These are not a charity effort. The desire is hooking companies on the drug forever. This is why there is the push to suggest kids not go to college. If a new generation is not trained to compete with AI then companies have no choice but to pay the insane costs of AI in the future. AI is a great tool but those pushing it hard may not have the best intentions of keeping it affordable.

For entry level positions AI isxa big threat because their tasks are typically the ones you use AI for when you're already an expert or a manager. But how to become an expert if you don't start learning the job at an entry level after graduation... ?

Important perspective. The conversation around AI is no longer just about capability. It's also about values, responsibility and the kind of future we want to build alongside these technologies.

A very valid point to consider while developing the AI governance framework.

The interesting question isn't whether the Pope is right or wrong. Most people already agree that AI should serve humanity. The hard problem is operational: how do we build systems that preserve human dignity when economic incentives, geopolitical competition, and institutional self-interest all push in different directions? The challenge is not defining values. The challenge is creating governance structures that remain aligned when those values become expensive. There is an even deeper observation: Nearly everyone in the thread assumes the problem is values. The harder problem is interpretation. Nobody says: "I oppose human dignity." Everyone says they support: dignity, justice, freedom, truth, accountability. The disagreement starts when those words must be operationalized. That's why governance is difficult. The battle is rarely over values. The battle is over what those values mean in practice. That is the most interesting thing in the thread, and almost nobody is talking about it.

Matthew Kilkenny This really resonates. What stands out to me is how much of this plays out through attention. If AI is becoming invisible moral infrastructure, then it is also shaping what we notice, what we value, and what we give our attention to over time. In an always-on environment, that influence becomes continuous. Attention is being guided long before we consciously reflect on it. So the question may not only be what AI is doing to the human spirit, but whether we understand attention well enough to protect it. Because that is where experience, judgment, and meaning are actually formed in real time. - James

I agree. Most students are not trying to break the rules, they are trying to navigate a system that often hasn't defined them clearly. The question is no longer whether AI should be used in higher education. Students are already using it. The real question is when, how, and for what purpose it should be used. Clear expectations, AI literacy, and assessment redesign are far more valuable than blanket bans. When institutions provide guidance instead of ambiguity, students can focus on learning rather than guessing where the boundaries are. The challenge isn't AI. It's governance, transparency, and intentional design.

When you push an update to an infrastructure tool utilized by billions, you do not treat the production environment like an A/B test sandbox. You provide detailed release notes. We are building high-level frameworks (like Continuous State Architectures and localized Omni-Sync nodes) on top of this API. When the frontend interface breaks silently, it fractures workflow momentum. Build the most powerful engine on the planet, but please, stop forgetting to install a functional door handle on the way out of the factory. #GoogleIO2026 #DeepMind #Gemini #UXDesign #SoftwareEngineering #AI #ProductManagement #TechUpdates

🚨 Silent UI Fragmentation: The legacy, single-tap TTS "Speaker" integration was quietly deprecated, forcing users to hunt for secondary "Read Aloud" workarounds just to access the new acoustic models. 🚨 Model Lock-In: The UI selector is currently glitching, locking power users into specific model tiers (Pro) without the ability to dynamically switch to Flash for lower-latency agentic workflows. 🚨 Compute Quotas over Message Limits: Shifting the governor to "Compute-Used" metrics without transparent dashboard tracking throttles developers running continuous-state logic or deep context windows.