Authority Sources

Subject: Forget the Gurus: Where the Real AI Investment Authority Lies

Hey Ins7ghts Crew,

Thinking about AI investment? Stop reading those breathless VC blog posts. The real authorities aren't hyping the next shiny object; they're the ones quietly building the infrastructure that makes AI possible. This week's data screams that truth.

Let's be blunt: the "thought leaders" are mostly noise. The signal is in the adoption and growth of foundational technologies.

Here's the breakdown:

My Hot Take: We're entering the "AI plumbing" era. The hype cycle is peaking. The real winners will be the companies building the reliable, scalable, and accessible infrastructure that allows everyone to leverage AI, not just the elite few.

Advice for You:

1. Look beyond the headlines. Don't be swayed by the latest AI unicorn. Focus on the underlying technologies that are enabling AI adoption.
2. Invest in data literacy. Train your teams to understand and interpret AI-generated data. Tableau (or similar tools) should be mandatory.
3. Prioritize infrastructure. Ensure you have the data pipelines, storage, and processing power to support your AI initiatives. Kafka, Spark, and Iceberg are good places to start.
4. Don't underestimate the power of existing tools. Find ways to integrate AI into your existing workflows, rather than trying to replace them entirely. Excel is your friend (for now).

Until next time, keep digging.

Yves Mulkers
The Ins7ghts Newsletter

Subject: Iceberg vs. Arrow: The Cold War Heating Up in Data Land

Forget AI hype for a minute. The real battle brewing isn't about models, it's about data. Specifically, how we store, access, and move it. This week's data deluge reveals a clear, albeit subtle, power struggle between two Apache behemoths: Iceberg and Arrow.

The affiliations are telling. While the articles mention individuals contributing to both projects, the context is crucial. We see Anurag Mantripragada, Anton Okolnychyi, Kevin Liu, and Russell Spitzer all tied to Apache Iceberg. Antoine Pitrou gets a nod for his work on Apache Arrow. But this isn't about individuals; it's about ecosystems.

Here's the provocative insight: Iceberg is positioning itself as the future of the data lake, while Arrow is quietly becoming the lingua franca for in-memory data processing. They're not directly competing yet, but their trajectories are on a collision course.

Why is this important? Because the winner (or the alliance that forms) will dictate the architecture of data platforms for the next decade. Think about it:

The noise? Endless articles about "modern data stacks" that fail to acknowledge this fundamental tension. Companies are blindly adopting both technologies without understanding the long-term implications. They're building systems that might work today, but could become architectural nightmares tomorrow.

My hot take? Iceberg needs to embrace Arrow more aggressively. Arrow's columnar memory format is a natural fit for Iceberg's data lake ambitions. Imagine querying petabytes of data in your Iceberg lake with Arrow's blazing-fast in-memory processing. That's the future. If Iceberg doesn't fully integrate with Arrow, it risks becoming just another table format in a crowded field.

Specific Advice:

1. If you're building a data lake, don't just pick a table format. Understand the implications for your entire data ecosystem. Evaluate how well it integrates with your existing tools and how it will support future use cases.
2. Invest in understanding Apache Arrow's Flight RPC framework. It's the key to unlocking zero-copy data transfer between systems and will be crucial for building high-performance data pipelines.
3. Follow the work of the Apache Iceberg and Arrow communities closely. Pay attention to their roadmap and how they are addressing the challenges of interoperability.

Don't be a passive observer in this data cold war. Choose your side (or, better yet, help forge an alliance) and build a data platform that's ready for the future.

Subject: Iceberg vs. Arrow: The Cold War Heating Up in Data Land (Part 2)

Forget AI hype for a minute. The real battle brewing isn't about models, it's about data. Specifically, how we store, access, and monetize it. Last week, we talked about ignoring the VC cheerleaders and focusing on the operators actually deploying AI. This week, let's dive into what they're saying – and what that really means.

The narrative being spun is all about "AI computing powers changing the stock market." (Source: Investing.com). We're seeing headlines touting portfolio performance, like "Year to date, 2 out of 3 global portfolios are beating their benchmark indexes, with 88% in the green." (Source: Investing.com). And the examples are always the same: "Our flagship Tech Titans strategy doubled the S&P 500 within 18 months, including notable winners like Super Micro Computer (+185%) and AppLovin (+157%)." (Source: Investing.com).

Sounds great, right? Wrong.

This is the arrow – the visible, sexy part of the AI investment story. Everyone's chasing the next Super Micro Computer. But underneath the surface, there's a massive iceberg of infrastructure, data management, and regulatory hurdles that will determine who really wins.

Think about it: Super Micro Computer's success is predicated on the demand for high-performance computing driven by AI. But who's building the data centers to house those servers? Who's providing the clean, labeled data to train those models? And who's navigating the increasingly complex data privacy regulations that govern its use?

Here's where the real money is being made, and where the long-term competitive advantage lies:

My Hot Take: The "AI computing powers" narrative is a distraction. It's shiny and exciting, but it obscures the fundamental challenges of data management. The companies that control the data will ultimately control the AI revolution. We're already seeing a shift from model-centric AI to data-centric AI, and this trend will only accelerate.

Your Action Items:

1. Audit your data supply chain: Understand where your data comes from, how it's processed, and who has access to it.
2. Invest in data governance: Implement policies and tools to ensure data quality, privacy, and compliance.
3. Focus on data literacy: Train your team to understand the value of data and how to use it effectively.

Stop chasing the arrow. Start building the iceberg. Your future depends on it.

Until next week,

Yves Mulkers
The Ins7ghts Newsletter