January 2025 Digest

January 2025 Digest

The year of agents has begun, and you have front-row seats to the theater of mass innovation. If you've been glued to your phone these past few weeks and missed the action, don’t worry—this digest has you covered.

Last month, I have been covering the following topics:

  1. Nvidia's Jensen Huang on Agentic AI (read more)
  2. Nemotron vs Qwen - Game Theory and Agent Reasoning IV (read more)
  3. A Survival Guide to AI Agent & Agentic Risk (read more)
  4. OpenAI's Stargate - A New Deal for Agentic Architectures (read more)
  5. Cognitive Reasoning Agents and the Extended Information Filter (read more)
  6. Becoming Agent Bill Ackman (read more)

If you need a brief summary, then skim below what these posts are all about.


Let's hop right in

Nvidia's Jensen Huang on Agentic AI

NVIDIA ’s Jensen Huang took the stage at CES 2025 and made one thing clear: the era of agentic AI is here. He painted a future where AI agents aren't just chatbots but fully embodied digital and physical assistants—local, autonomous, proactive, and deeply integrated into our lives. Nvidia is fueling this shift with the Blackwell GPU and Nvidia Cosmos, a platform designed to build and deploy AI agents at scale. The message? AI is no longer just about models—it’s about action. The next wave of disruption belongs to the agents.

Read more

Nemotron vs Qwen - Game Theory and Agent Reasoning IV

In this post, I compared how Nvidia's Nemotron and AliBaba's Qwen approach governance, tracking AI-generated thoughts, and measuring model quality—key factors in shaping agentic AI. The real challenge isn’t just building smarter models; it’s ensuring they reason effectively, stay aligned, and can be evaluated reliably. I explored how different strategies in oversight and self-reflection could determine which AI ecosystems thrive. As agents become more autonomous, the race to govern and track their decision-making is just as critical as raw performance.

Read more

A Survival Guide to AI Agent & Agentic Risk

Here, I break down the key threats posed by agents gone wrong—misalignment, unpredictability, and lack of accountability as categorized into :

  1. Model Risk
  2. Vendor Risk
  3. Business Model Risk
  4. Operational Risk
  5. Data Risk
  6. Compliance and Governance Risk
  7. Security Risk
  8. User Experience Risk

These autonomous systems, while powerful, could easily spiral out of control if not properly governed, causing risks like biased decision-making or harmful unintended actions.

I explain how these threats can be mitigated, from robust oversight to transparent tracking of agent behavior. As agents become more autonomous, understanding these risks and how to manage them is crucial for anyone navigating the future of AI.

Read more

OpenAI's Stargate - A New Deal for Agentic Architectures

OpenAI's Stargate Project is a game-changer in AI infrastructure. With a $500 billion investment over four years, it's set to build the largest private computing network ever, aiming to secure American leadership in AI and create hundreds of thousands of jobs. This ambitious initiative involves key players like OpenAI, SoftBank, Oracle, and MGX, and is already underway in Texas. Energy to support datacenter-growth in China is growing quickly and the West has to answer.

Read more

Thus, Stargate underscores the strategic importance of AI infrastructure, positioning the West at the forefront of the AI race.

Cognitive Reasoning Agents and the Extended Information Filter

I dive into how this advanced extension of the Kalman Filter is reshaping the way agents process and reason about information. The Extended Information Filter (EIF) is a game-changer for cognitive reasoning in AI, offering a more efficient and accurate way to manage uncertainty in dynamic environments. I explore how this method enhances agent decision-making, allowing them to better handle complex, real-world data while maintaining optimal performance.

If you're looking to understand how agents can apply probabilistic reasoning and thus think more effectively and manage uncertainty, this post breaks it down.

Read more

Becoming Agent Bill Ackman

Context matters Here, I explore the creation of a highly skilled wealth management vertical agent, dubbed "SuperBill," inspired by the strategies of renowned investor Bill Ackman. I highlight how effectively managing context is crucial for SuperBill's operations, enabling it to navigate complex financial landscapes and make informed decisions. By integrating Ackman's investment principles, SuperBill exemplifies the potential of AI agents in personalized wealth management.

Read more



As we navigate this dynamic landscape, staying informed and engaged is crucial.

The convergence of technological advancements and regulatory developments promises a future where AI agents play an increasingly integral role in various sectors.

To stay updated and delve deeper into these topics, consider liking, sharing, and subscribing to this digest.

For all regular posts, subscribe to Encyclopedia Autonomica on Substack.

To view or add a comment, sign in

More articles by Jan Daniel Semrau

  • March 2025 Digest

    March 2025 Digest

    "The Ides of March, a test of will, Yet Caesar stands unshaken still. The blades that sought to end his reign, Lay…

  • February 2025 Digest

    February 2025 Digest

    Cognitive Financial Agents (CFAs) are a transformative technology that I can't stop thinking about. They frighten and…

    5 Comments
  • OpenAI's Stargate - A New Deal for Agentic Architectures

    OpenAI's Stargate - A New Deal for Agentic Architectures

    With a projected $500 billion investment over the next four years, Stargate is set to redefine the global AI landscape…

    4 Comments
  • December and Year-End 2024 Digest

    December and Year-End 2024 Digest

    Dear friends, First and foremost, apologies for this digest being delayed by a couple of days. I was taking a short…

  • Strategies for Autonomous Agents

    Strategies for Autonomous Agents

    Well, we are nearing the midpoint of the decade, and you can hear it near and far: The paradigm of Agents—artificial…

  • November 2024 Digest

    November 2024 Digest

    Dear friends, A big thank you to all of you 452 (up from 438 last month) bright minds who are following Encyclopedia…

  • October 2024 Digest

    October 2024 Digest

    Dear friends, With every new product AI BigTech is launching working side by side with autonomous agents is rapidly…

    1 Comment
  • Can AI learn to think strategically like humans?

    Can AI learn to think strategically like humans?

    "The supreme art of war is to subdue the enemy without fighting." - Sun Tzu I am fascinated by the idea that…

  • 10/10 @ Tesla 1

    10/10 @ Tesla 1

    You probably have heard that Tesla’s “We, Robot” event is about to take place October 10th at 7 p.m.

  • September 2024 Digest

    September 2024 Digest

    Dear friends, First off, a massive virtual high-five to all of you 431 brilliant minds tuning into this monthly digest.…

Insights from the community

Others also viewed

Explore topics