top of page

Another Crazy Day in AI: Agents, Algorithms, and American Strategy

Another Crazy Day in AI: An Almost Daily Newsletter

Hello, AI Enthusiasts.


How do you keep up when everything is moving at once?


A new podcast brings together a week of very different headlines: math medals, autonomous agents, and policy blueprints. One thread? Progress is messy—and fast.


Over in the boardroom, bold AI investments are falling flat, but not for lack of tech. Turns out, change fatigue is real, and resilience might be the missing skill.


And ChatGPT’s new “study mode” wants to teach students how to learn, not just copy and paste. Maybe AI is finally doing its homework?


Call it chaos or progress. Either way, it’s happening.


Here's another crazy day in AI:

  • Technical breakthroughs and infrastructure realities

  • How to build change resilience for AI

  • OpenAI debuts new learning mode in ChatGPT

  • Some AI tools to try out


TODAY'S FEATURED ITEM: Technical Milestones and Public Planning

A robotic scientist in a classic white coat with 'AI Scientist' on its back stands beside a human scientist with 'Human Scientist' on their coat, looking towards the AI Scientist.

Image Credit: Wowza (created with Ideogram)


What should progress in AI actually look like... and who decides when it’s meaningful?



This week’s episode of Mixture of Experts brings together a full spectrum of those developments, featuring DeepMind’s showing at the International Math Olympiad, OpenAI’s agent rollout, and the White House’s newly released AI Action Plan.


Hosted by Tim Hwang, the discussion includes Kate Soule, Gabe Goodhart, Mihai Criveti, and guest Ryan Hagemann, unpacking where things stand at the intersection of AI capability, usability, and regulation. It’s not one story, but several threads that signal how much is happening across technical benchmarks, practical tools, and national frameworks for governance.



Main Points Discussed:

  • DeepMind and OpenAI systems reached gold-level performance at the International Math Olympiad, placing them among the top 10% of high school mathematicians globally

  • Today's AI problem-solving involves orchestrating millions of tokens across specialized tools and verification systems, a far cry from simple prompt-and-response interactions

  • OpenAI's ChatGPT agents enable users to start tasks and walk away, returning later to completed work, though security concerns limit enterprise readiness

  • The ecosystem of AI tools and protocols remains fragmented, creating demand for middleware solutions like the MCP Gateway project to handle authentication and standardization

  • The White House AI Action Plan endorses open-source development and outlines infrastructure expansion while promoting American technology exports

  • As AI tackles increasingly specialized domains, evaluating performance becomes more expensive and relies on smaller pools of expert human judges

  • Consumer applications continue advancing while enterprise deployment lags due to trust, security, and credential management challenges



The mathematical breakthrough deserves attention, but perhaps not for the reasons you might expect. Yes, solving International Math Olympiad problems represents genuine reasoning capability rather than pattern recognition. However, as the panelists noted, this feels more like AlphaGo than a fundamental turning point. The achievement demonstrates sophisticated tool use and parallel processing techniques that cost significantly more than traditional AI approaches—potentially thousands of dollars per complex problem solved.


What makes this discussion particularly valuable is how it illustrates the gap between what AI can do and what we're comfortable letting it do. The same systems that can tackle graduate-level mathematics still face basic hurdles around trust and security for everyday business tasks. Meanwhile, policymakers are working to balance innovation with infrastructure needs, recently endorsing open-source development after months of uncertainty. The conversation reveals an ecosystem where technical capability often runs ahead of practical deployment, raising questions about how we measure progress when the most impressive benchmarks don't necessarily translate to immediate real-world impact.



Watch it on YouTube here.

Listen on Apple Podcasts here.

Listen on Spotify here.

OTHER INTERESTING AI HIGHLIGHTS:


How to Build Change Resilience for AI

/Karim R. Lakhani, Jen Stave, Douglas Ng, and Daniel Martines, on Harvard Business Review


Despite bold AI investments, most organizations are falling short of meaningful results—not because of technology, but because of people. This piece from Harvard Business Review introduces “change resilience” as the core capability leaders need to navigate AI’s fast-paced disruptions. By building three core muscles—sensing, rewiring, and lock-in—companies like Shopify, Accenture, and P&G are showing how to turn constant change into lasting value. The article also includes a five-step playbook to help organizations shift from reactive transformation to a sustained, learning-focused strategy.



Read more here.


OpenAI Debuts New Learning Mode in ChatGPT

/OpenAI


OpenAI is rolling out “study mode” in ChatGPT—a new learning experience designed to help students work through problems instead of just getting quick answers. With features like scaffolded responses, interactive prompts, and built-in knowledge checks, study mode supports deeper understanding and active learning. It's available to Free, Plus, Pro, and Team users starting today, with ChatGPT Edu support coming soon. Designed with the help of educators and grounded in learning science, study mode reflects a shift in how AI can enhance—not shortcut—education.



Read more here.

SOME AI TOOLS TO TRY OUT:


  • PopResume – Create resumes, cover letters, and prep for interviews with AI.

  • Seed LiveInterpret 2.0 – Translate voice conversations while preserving your voice.

  • Tapflow – Turn your documents into sellable guides, playbooks, and workflows.


That’s a wrap on today’s Almost Daily craziness.


Catch us almost every day—almost! 😉

EXCITING NEWS:

The Another Crazy Day in AI newsletter is on LinkedIn!!!



Wowza, Inc.

Leveraging AI for Enhanced Content: As part of our commitment to exploring new technologies, we used AI to help curate and refine our newsletters. This enriches our content and keeps us at the forefront of digital innovation, ensuring you stay informed with the latest trends and developments.





Comments


Subscribe to Another Crazy Day in AI​

Catch us almost every day—almost! 😉

Thanks for signing up!

Copyright Wowza, inc 2025
bottom of page