Rabih Nassar

In an AI-enabled world, how our systems connect is becoming as crucial as what they connect. This insightful post from Glama.ai outlines why MCP (Model Context Protocol) matters—and why the old API paradigms may no longer suffice for agentic models. Instead of brittle REST endpoints and scattered schemas, MCP offers a unified, discovery-based, bidirectional protocol built for AI-first interoperability. It’s like moving from static documentation to GPS navigation for autonomous systems. Curious how this affects AI integration at scale? https://lnkd.in/gK9wB2gm #AIFoundation #AIInteroperability #MCP

14

2 Comments

Philip Abraham ~ Chief Innovation Officer and Strategist Brilliancy Deep Tech The Solution: #Unified #Intelligence #Architecture- #Complexity #Science The #Brilliancy Chain™ transforms this landscape. It provides a decentralized, #complexity-#science-based framework that integrates all systems under a single verifiable logic layer. Trey Rutledge Nino Marcantonio Jaime Campos Flores Mark A. Mona

3

1 Comment

🔹 Empire of AI: OpenAI, Economic Power, and National Security 📍 Hosted by: Stanford University Graduate School of Business – Corporations and Society Initiative (CASI) & Program on Capitalism and Democracy (CAD) A recent session at Stanford GSB examined the impact of artificial intelligence on global capitalism and national security. The event brought together Evan Epstein (Lecturer, Stanford GSB) and Karen Hao (investigative journalist and author of Empire of AI) to discuss the implications of capital concentration, governance models, and strategic investments within the AI ecosystem. Key insights highlighted that the rise of OpenAI has not only intensified competition among technology firms but has also triggered new forms of capital concentration. Epstein emphasized that corporate ROI and efficiency metrics intersect with national security considerations, while Hao drew attention to the weakening of democratic accountability mechanisms in this process. Available data demonstrate that AI investments exhibit high economic return potential while also creating unprecedented levels of capital concentration. According to Crunchbase, approximately 45% of global AI funding in Q2 2025 was directed to only 16 companies, with mega rounds such as those by Scale AI standing out. Similarly, AI startups captured 57% of global venture capital investment in Q1 2025. EY data further reveal that when organizations allocate at least 5% of total budgets to AI investments, measurable ROI increases of 75–84% are observed in operational efficiency, cybersecurity, customer satisfaction, and product innovation. This dual reality—high ROI potential coupled with concentration-related risks—underscores the necessity of establishing institutional frameworks that balance economic efficiency with societal and market resilience. Hence, a sustainable AI ecosystem must be built not solely on financial performance metrics but also reinforced through systemic resilience, transparency, and regulatory capacity.

6

My agency, Mostec.ai, is a management-led AI company for a reason. MIT says 95% of AI projects generate zero return. BCG adds that “70% of AI transformation is people.” That matches what I’ve seen. Here is a simple example. A company struggling with lead generation installs an AI automation system that suddenly brings in 3–4x more leads. Not only do they have enough for every salesperson, but they now have twice as many as they can comfortably handle. Excitement is high, salespeople push hard… until it becomes overwhelming. Almost inevitably, instead of taking responsibility for every lead, they begin to select, picking the “best” leads and ignoring the rest. The logic sounds reasonable. Rather than giving quick, shallow attention to all leads, they picked only the most potent ones and invested more energy there. Seems smart, right? Focus energy where it matters most. Except it rarely works. When I ran the top outbound insurance sales teams, agents did the same thing. They would look at the lead data and decide to call only the ones that looked the most promising. For example, city residents instead of rural ones. Certain profiles, instead of others. Those who already have the most insurance policies. Again and again, the results showed the same: it was far better to define solid qualification criteria upfront and then systematically process every lead. This kind of situation has very little to do with AI itself. In both cases—whether employees follow through on every lead or start cherry-picking—the AI works perfectly. The difference is in management. Technology today does what it’s supposed to do. The real results are created where AI stops, and human nature takes over. That’s why at Mostec, we center our AI systems around people, not just tools. It’s people that make the difference. Always. --- In your experience, where do AI projects stumble more: the technology, or the management around it?

10

2 Comments