In episode 20 of Open Source Ready, Brian and John sit down with data engineering and cognitive science expert Vasilije Markovic to explore AI memory and how we can build more intelligent systems. From the challenges of "context rot" to the practical applications of AI memory in construction, education, and finance, this conversation covers how to give your AI the context it truly needs.

Subscribe

Open Source Ready - Ep. #20, Exploring AI Memory with Vasilije Markovic of Cognee

This week’s guest is Bill Murphy, managing partner at Cresting Wave 
0:00 – Tony and Shen discuss System of a Down and Gorillaz 
6:00 – Bill Murphy, Blackstone’s former CTO, joins to discuss that much-discussed MIT study on AI projects failing and factors executives should consider as the technology continues to evolves

Episode 331: Cresting Wave’s Bill Murphy

Welcome to "AI with Shaily," hosted by Shailendra Kumar, a knowledgeable and approachable guide in the world of artificial intelligence 🤖✨. Shaily breaks down complex AI topics with clarity and insight, making them accessible without confusing jargon—just straight talk.

In this episode, Shaily dives into the hot topic of the AI community: the difference between AGI (Artificial General Intelligence) and ASI (Artificial Superintelligence) 🧠⚡. He explains that while AGI is AI that thinks and learns like a human across a wide range of tasks—imagine an AI that can seamlessly switch from cooking recipes to solving math puzzles—ASI goes way beyond that. ASI is like a supercharged genius that not only outperforms humans in every domain but can invent new sciences, rewrite the rules, and improve itself endlessly 🔥🚀.

Shaily shares a handy comparison table to highlight key distinctions:

- Scope: AGI matches human-level intelligence broadly; ASI far surpasses human intelligence in all areas.
- Learning: AGI learns and generalizes like humans; ASI self-improves recursively without limits.
- Capabilities: AGI reasons and adapts; ASI invents new science, creative works, and strategies.
- Self-awareness: AGI understands its responsibilities; ASI may develop complex self-awareness and even emotions.
- Impact: AGI is a major milestone; ASI could transform civilization itself.
- Current Status: AGI is near, with experts like Sam Altman of OpenAI optimistic about early signs; ASI remains a distant goal requiring breakthroughs.

Shaily emphasizes why understanding this difference matters: AGI will change how we interact with machines, turning them into partners in innovation, while ASI raises profound ethical and societal questions about control and the future of humanity 🌍🤝.

He shares a personal story reflecting on his early fascination with AI and how witnessing AGI traits emerge today feels like watching a toddler take its first steps—exciting but just the beginning of a long journey toward ASI, which remains full of unknowns and challenges 👶➡️🌟.

For listeners eager to stay ahead, Shaily advises not only keeping up with technical advances but also engaging deeply with ethical discussions and critical thinking about AI’s societal role. This balanced approach is key to navigating the evolving AI landscape 🧭💡.

He closes with a thoughtful quote from Alan Turing about seeing just a short distance ahead but recognizing much work to be done—AGI is that short distance, while ASI represents the vast horizon beyond 🌅.

Shaily invites the audience to connect with him on YouTube, Twitter, LinkedIn, and Medium for more AI insights and encourages subscribing, commenting, and sharing thoughts about what AGI and ASI might mean personally and globally.

With warmth and encouragement, Shailendra Kumar signs off reminding everyone to keep questioning, learning, and imagining the future together 🌟🤗.

AGI vs ASI: The Future of Artificial Intelligence Unveiled

In our latest episode, Barry Render sits down with David Rosenthal, a recent MBA graduate from the University of Texas at Austin, to explore how artificial intelligence is transforming business education and early career experiences.

David shares insights into AI in the classroom, AI as a central part of his internship, and his entrepreneurial journey using AI to build Fantasy Fusion Sports.

OM Podcast #40: AI in the MBA Classroom and Beyond

CFIGS Research Spotlight - Episode 2 (Mapping Tomorrow with Geospatial Intelligence)

What if the next leap in AI isn’t bigger at all? Samsung just unveiled a tiny model built to run on your phone, earbuds, and watch.
We unpack why on-device AI matters: faster responses, lower costs, and stronger privacy with your photos, voice, and messages processed locally. Harry Kermode explains how tiny AI works: distillation, pruning, quantization, and NPUs that wake only when needed. We also map the trade-offs and the likely hybrid handoff between edge and cloud.
Expect everyday devices to feel smarter by default, reshaping app design, battery expectations, and how personal your AI becomes without sending everything to the cloud. Plus, the privacy fine print and what benchmarks and tools to watch next.
If you enjoyed this explainer, follow AI Daily, share the episode, and leave a review. 

FULL TRANSCRIPT: I can’t replicate that show’s exact style, but here’s a concise, narrative news explainer in a similar tone.

From AI Daily, I’m Harry Kermode.

Today: what if the next big leap in artificial intelligence isn’t bigger at all? Samsung has unveiled a tiny AI model designed to run on the devices we carry—phones, earbuds, even watches. It’s a bet that the future of AI lives at the edge, not in the cloud.

Here’s why that matters. When AI runs on your device, your data doesn’t need to travel. That means faster responses, lower costs, and a real shift in privacy—your voice, your photos, your messages can be processed locally, not on someone else’s server.

Imagine your morning: you dictate a text on the subway, offline. Your watch translates a phrase on the fly. Your camera de-blurs a photo as you snap it. No spinning “connecting” icon. No signal? No problem.

Under the hood, “tiny AI” isn’t magic—it’s engineering. Start with a capable model, then compress it: distillation to teach a smaller network the tricks of a larger one, pruning to snip away redundant connections, quantization to store numbers more efficiently. Pair that with modern phone chips—neural processing units tuned for matrix math—and clever scheduling so models wake only when needed. The result: meaningful intelligence within a tight power budget.

There are trade-offs. Smaller models can’t juggle sprawling tasks or remember long conversations. Expect a hybrid approach: on-device for speed and privacy, the cloud for heavy lifting, with seamless handoffs you barely notice.

The stakes are big. If everyday devices become smart by default, AI stops being a destination and becomes an atmosphere. That’s democratizing—and disruptive. App makers will design for on-device first. Carriers could see fewer data-hungry requests. Companies will compete on battery life, latency, and how well tiny models adapt to you personally.

And privacy? On-device processing keeps data local, but it isn’t a free pass. Companies will need transparent update policies, secure enclaves for sensitive data, and clear controls so you can decide what stays on your phone and what leaves it.

What to watch next: real-world benchmarks on speed and battery drain, developer tools that make building tiny-model apps easy, and whether Samsung’s move pushes rivals to standardize how edge AI is measured, updated, and secured.

Small models, everywhere. That might be how AI finally becomes part of the background—powerful, personal, and almost invisible.

That’s AI Daily. I’m Harry Kermode. See you tomorrow.

The Future of AI Is Small and On-Device

In this special holiday-weekend episode, Brian interviews Gary Ross, Associate Professor in the Faculty of Pharmaceutical Sciences at Kanazawa University, where he is Director of English and develops the online ESL program. Gary is a full-stack developer specializing in AI and online learning. He is the creator of J-Talk.com and the superb conference platform Edzilla, which we continue to use as our JALTCALL conference hosting system. We discuss Gary's various Kaken projects, including (1) Speech recognition in medical communication and training using ML with Jeanette Dennison, and (2) the Extensive Reading Artificial Intelligence app in collaboration with Mark Breirley https://gamma.erai.app/dashboard/manuscripts 
Check out Gary's J-Talk website for academics working in Japan(ese). https://j-talk.com/convert
and contact us at the podcast using jaltcall@jaltcall.org 
Contact Gary Ross at gary@p.kanazawa-u.ac.jp Kaken Researcher Number 10708142

Gary Ross Interview by Brian Gallagher Oct 2025

Choosing between a Data Science bootcamp and a traditional degree can shape your tech career path. This blog explores the key differences in duration, cost, curriculum, and job readiness to help you decide the right learning journey. Understand which route best suits your goals and future in data science.

Data Science Bootcamps vs. Traditional Degree Programs Which is Right for You

A structured, systems-level exploration of reflection and recursive development in intelligent systems.

This Technical Perspective from Chapter 22 analyses how sustaining intention and growth in AI requires memory mechanisms, calibration thresholds, and recursive self-assessment.

Drawing parallels with biological cognition and metacognitive learning, it explains how loops may decay or evolve depending on system design.

Recorded in a measured technical tone (Audacity, Arty-style), this voice track supports deeper engagement with the conceptual architecture of spiral evolution.

chapter 22 section 4 technical perspective

In this conversation, Dave Prior and Hugo Bowne-Anderson discuss the evolving landscape of AI and data science, focusing on the role of AI agents in solving business problems. Hugo shares insights on how to effectively implement AI solutions, the importance of understanding the underlying data, and the need for continuous improvement in AI systems. They also touch on the skills necessary for navigating the AI landscape, the value of collaboration between technical and non-technical teams, and the importance of assessing the value of AI projects. Hugo concludes by offering a course on building AI applications, emphasizing the iterative nature of AI development.

Takeaways

 - Hugo emphasizes the importance of data in AI applications.
 - AI agents can automate tasks but require human oversight.
 - Understanding the problem is crucial before implementing AI solutions.
 - Prompt engineering remains a valuable skill alongside learning about agents.
 - Consultants should educate clients on practical AI applications.
 - AI systems should be built incrementally and iteratively.
 - Value assessment in AI projects should focus on efficiency and cost savings.
 - Continuous improvement is essential for AI systems to remain effective.
 - Experimentation with AI tools can lead to innovative solutions.
 - Collaboration between technical and non-technical teams is vital for successful AI implementation.

Chapters

00:00 Introduction to Data and AI Literacy
06:14 Understanding AI Agents vs. LLMs
09:18 The Role of Agents in Business Solutions
12:21 Navigating the Future of AI and Agents
15:24 Consulting and Client Education in AI
18:37 Building Incremental AI Solutions
21:29 The Future of AI Coding and Debugging
24:32 Prototyping with AI: Challenges and Solutions
25:32 Leveraging AI for User Insights and Competitive Analysis
27:29 Understanding Value in AI Development
32:05 The Role of Product Managers in AI Integration
33:00 AI as an Instrument: The Human Element
35:33 Getting Started with AI: Practical Steps for Teams
38:51 Building AI Applications: Course Overview and Insights

Links from the Podcast:

Stop Building AI Agents - Here’s what you should build instead (Article) https://www.decodingai.com/p/stop-building-ai-agents
Anthropic https://www.anthropic.com/engineering/multi-agent-research-system
The Colgate Study https://www.pymc-labs.com/blog-posts/AI-based-Customer-Research

Hugo’s Course (Starts November 3, 2025)
Building AI Applications for Data Scientists and Software Engineers (with a 25% discount)
https://maven.com/hugo-stefan/building-ai-apps-ds-and-swe-from-first-principles?promoCode=drunkenpm
(You can use the discount code drunkenpm to get 25% off)

How To Be A Podcast Guest with Jay Hrcsko https://youtu.be/vkNbgwcolIM

Contacting Hugo
LinkedIn https://www.linkedin.com/in/hugo-bowne-anderson-045939a5/
Substack https://hugobowne.substack.com/

Contacting Dave
Linktree: https://linktr.ee/mrsungo
Dave’s Classes: https://www.eventbrite.com/cc/dave-prior-classes-4758623

Navigating AI - Making Sense of Agents and When To Use Them with Hugo Bowne-Anderson

Heavybit is the leading early-stage investor in enterprise infrastructure. Since 2013, we've helped launch and scale visionary technical startups—from DevSecOps and feature flagging to AI code generation and beyond. We go hands-on from day 0, backing founders as they turn code into companies and build the future of software from the bottom up to the top down.

Unlock the secrets of data mastery with Shailendra, a 27-year veteran in AI, ML, and Data Science, driving AU$ 6 Billion in transformative value. As the author of 'Making Money out of Data,' their strategies redefine digital success. With expertise honed at IBM and Accenture, they drive innovation and data monetization. Explore how loyalty programs like Flybuys and Everyday Rewards revolutionize business—connect with Shaily today.

Japan association for Language Teaching (JALT) Computer Assisted Language Learning Special Interest Group. (JALTCALL)

A podcast about Agile and Project Management

Gradious is an EdTech platform dedicated to empowering students, graduates, and professionals with industry-ready skills. Our mission is to bridge the gap between academic learning and real-world requirements by offering hands-on training, coding bootcamps, and upskilling programs that prepare learners to excel from day one in their careers.

Wei-Shen Wong and Anthony Malakian discuss the hottest financial technology topics in the capital markets every week.

New tracks tagged #machine learning