Browse Comments — Relevant (AI ∩ value)

This points to a deeper shift than cost curves. What’s breaking isn’t just the price of inference, it’s the assumption that intelligence must live inside monoliths. When scaling hits diminishing returns, architecture becomes the lever. Ensembles move the bottleneck from raw compute to orchestration, signal interpretation, and aggregation logic. That’s a fundamentally different design philosophy, and it aligns much more closely with how real-world intelligence actually works. If this holds, the next frontier isn’t bigger models competing with OpenAI or DeepSeek AI on brute force. It’s smaller, purpose-built systems coordinated intelligently, closer to perception, context, and decision-making. The age of scaling was inevitable. An age of architecture was always next.

Mike Pappas How do you know that they "want individual AI-powered solutions to specific tasks"? Seems like a reach (unless people have told you that directly in an interview). I feel like it may be more accurate to say: 1. They want AI to cost less (so they aren't worried about driving their bill through the roof -> see Jackson Oaks) 2. Give them faster responses (so they don't sit there staring at it wondering what to do in the meantime) 3. While still accurately and actually solving their problem (e.g. doing the task) There's a reason I use Opus for almost everything: I trust the response will actually be good and be a solution to my problem: good email, working software, etc. Personally I don't care in any way shape or form about "individual AI-powered solutions to specific tasks" I just want my "task" (problem) done (solved) for cheap. xD

Nice automation. How do you prevent losing your authenticity and credibility with your followers if AI is doing all of this for you and you no longer personally contribute? And what value do social media platforms still have if AI agents start posting and replying?

Great job! Now I'm waiting for somebody to build an AI system that consumes LinkedIn posts, so we all don't have to watch all these AI-generated posts. Just imagine how much time that would free up, what efficiency!

Going from one sentence to a fully published video is wild. The fact that this runs on 17 specialized skills and handles everything autonomously shows how far agentic workflows have come. Great build.

Frankly speaking I'm really frustrated by the amount of AI-generated content on YT. It is kinda clickbait, you start watching, after a minute or two realizes it is AI and then you stop watching. Simply because of poor quality, glitches and generic inaccurate video of no value.

Most people don’t actually need “autonomous agents,” they need reliable content systems that don’t break when the input changes or the platform shifts. I see this a lot while building Collio AI

Enrique Marq, I really appreciate the engineering and the idea behind this, it’s impressive work. I do wonder, though: what’s the end goal when this level of automation is applied to social media, which originally existed for human expression and connection? If our social presence is generated by AI, posts written in our “voice,” scripts authored by systems trained to sound like us, what are we actually doing when we “connect” online? At that point, it feels less like people sharing genuine thoughts and reactions, and more like AI systems interacting with each other on our behalf. If everyone is doing this, what’s the point of socializing at all? Part of the value of connection is the time, effort, and intent behind it. Writing something yourself, crafting a thoughtful message, spending attention, that effort is a signal that the people you’re talking to matter. It’s a bit like a handwritten note or choosing a meaningful gift. The effort itself carries meaning. If we outsource that entirely, we may gain efficiency, but we lose the original purpose: the human care behind the act.

I looked at the comments section here to check how many people tolerate and appreciate being told what to do and how to do it without experimentation and data about real results. And I’ll say this again: Ai doesn’t belong on LinkedIn. 90% of automation connection with the LinkedIn app is being either PUNISHED by the algorithm or BANNING accounts. If not today, then “tomorrow”. Not to mention Ai is thinking for all of us, based on everything it learned about each one of us. And it is using average fractions of the internet. If you saw who was feeding Ai you wouldn’t use it for a shortcut

Hey man! We gotta talk. I have built a Distributed AI Infrastructure platform, and one of my platform’s core abilities is the ability to analyze media formats for enterprises to learn from with cross domain learning capabilities. We also do this, while ensuring our clients keep their data sovereign. I have an open-source version of my Distributed AI Infrastructure platform on GitHub (doesn’t come with the above capability, that is the full version - which is already built), but it comes with highly useful capabilities that you and your dev team can easily play around with and integrate into your platform for efficiency gains, intelligence expansion, and cost mitigation on compute/energy within your organization. Excited to talk about AI and orchestration as we are mutually aligned in that category!

Impressive level of automation, but the real question is elsewhere: where does human value sit in the loop? If everything becomes generation + distribution, differentiation won’t be production anymore but framing, ideas, and editorial intent. That’s where the human layer becomes critical again.

Maarten Masschelein agreed this hits where most teams fail. Data stewardship isn’t a title, it’s discipline. Clean pipelines mean nothing if people don’t trust the data. The real test: can someone use your data without asking you? If not, it’s not a tech gap, it’s an ownership gap. Get this right, and AI delivers. Get it wrong, and you just scale confusion.

In the context of AI, informal data stewards are the people catching the problems that models will eventually amplify. The person who documents dataset quirks before they become training data assumptions is doing governance work that no formal review process will surface in time. That behaviour has always mattered, but even more now.

Recognizing an Organizational Data Steward is a sign of organizational maturity. For this to work, a specific mindset must exist: "Stop fixing bad data; start tuning the process!" The shift from liability to strategic asset only happens when we stop treating data stewardship as a solo role and start seeing it as a collective responsibility. The twist? Everyone who manages a process is a Contributing Data Steward, led by Organizational Data Stewards. Now, for AI to be credible, the accountability must lie with those who feed the model.

The agent-reasoning and agentic_rag implementations in this repo are standout features. I'm currently deep in the trenches building an agentic RAG system for the AWS Well-Architected Framework, and moving past simple retrieval to 'Cognitive Architectures' (like ReAct or CoT) is where the real value is. It’s one thing to get an LLM to chat; it’s another to build a harness that ensures data integrity across 3,000+ chunks. Great share!

Hi Abhishek Veeramalla The repository appears highly valuable for accelerating practical AI engineering adoption. However, from enterprise security architecture, and AI governance perspective, this type of “production-ready AI” narrative must be evaluated very carefully because operational AI deployment risk is significantly more dangerous than experimental AI learning risk. The biggest concern is not whether the demos work. The real concern is whether developers unknowingly normalize insecure AI architecture patterns into enterprise production environments. Open-source AI acceleration without mandatory governance-by-design can create scalable technical debt, compliance exposure, and systemic AI security fragility at enterprise scale. What formal threat modeling methodology was used to validate the security posture of these “production-ready” AI architectures? What mechanisms ensure explainability, traceability, and auditability for autonomous reasoning frameworks such as ReAct, ToT, and CoT?

GitHub is no longer just a “code repository website” GitHub today is far beyond an AI learning platform. It has become a global engineering collaboration hub across all technologies — AI, DevOps, SAP, cloud, analytics, automation, and enterprise applications. The real value is not just code hosting, but how communities share reusable knowledge, accelerate innovation, and build production-grade solutions together. #GitHub #OpenSource #AI #DevOps #SoftwareEngineering #Innovation #Technology #CloudComputing #Automation #MachineLearning #SAP #EnterpriseTechnology #DigitalTransformation #Collaboration #Developers #CICD #DataEngineering #MLOps #ArtificialIntelligence #TechCommunity

This is the kind of resource that shortens the gap between consuming AI content and actually building with it. A lot of people are stuck in endless learning loops right now. Repositories like this become valuable because they move learning from theory into implementation, systems thinking, experimentation, and real-world problem solving

The #1 most used model on OpenRouter right now costs $0.07 per million tokens (Hy3 preview). The #2 costs $5.00 (Claude Opus 4.7). Wild gap. check out for more upto date information

You forgot to add Hyperlambda.dev The Best Solution for building AI Agents with every generated code said to be 100% mathematically correct. Being the only LLM AI Agent in the world with the most accurate fine-tuning model, it said to come at a fractional cost of 0.000001% of Claude AI. We should be expecting a tsunami of projects hitting the market just because a model got it right 👍