Welcome to the AI Pulse Podcast hosted by Dr. Marlene Wolfgruber and Maxime Vermeir of ABBYY, an intelligent automation company. This series covers a wide range of exciting topics, all related to the latest developments in AI, intelligent automation, and business.
We’ll bring on industry leaders to chat about their specific fields, what they’re seeing in the market, and their firsthand perspectives on how purpose-built AI is putting global enterprises’ information to work. Listen in and stay up to speed with industry experts!
What Generative AI Actually Changes, and Where Agents Take it Next feat. Alan Nichol, CTO of Rasa
AI didn't just get better at answering questions, it started acting on them. In this episode of ABBYY's AI Pulse Podcast, we sit down with Alan Nicol, co-founder and Chief Technology Officer (CTO) of Rasa, to unpack what generative AI has actually changed, where the hype outran the results, and what comes next as AI agents take on more autonomy.
Alan walks through the "jagged frontier" of AI capability, where some problems have been solved dramatically well and others remain mostly a matter of taste.
He breaks down what a "harness" actually is and why it matters more than the raw model, explains when to hand an agent judgment versus keeping a process rule-based, and shares why so many AI demos impress in minutes but struggle once they reach production. He closes with a look at where this is headed: agents that can validate their own work, and more of what we build defined in code.
---
Visit the ABBYY website to explore our platform and start driving higher accuracy today.
Contact us for a demo: https://bit.ly/Request-ABBYY-Demo (https://bit.ly/Request-ABBYY-Demo)
---
Subscribe to the AI Pulse Podcast for more conversations on how AI is reshaping the way businesses work.
---
Chapter Markers:
5:12 – Generative AI vs. Agentic AI: Definitions and misconceptions
15:07 – Agents and Rule-based systems: Where each fit, and how they can work together
22:14 – What is a "Harness"? Building structure around LLMs
26:52 – Why demo's fall apart in production
29:53 – The build-vs-buy shift
33:27 – Advice for listeners and what's next for agents
---
Have a question you'd like us to ask our next guest? Drop it in the comments! ---
#AIPulsePodcast #ABBYY #GenerativeAI #AIAgents #IntelligentAutomation
30 Jun 2026
Why Al Is Changing Who Gets to Build feat. Tom Shaw
TL;DR: Tune in to this episode with Tom Shaw for critiques on influencer-driven hype around autonomous agents like “OpenClonk,” security risks, “smoke and mirrors,” and to explore how AI will reshape the industry looking at SaaS value and pricing models. ---Software developer and content creator Tom Shaw joins us to share about his accidental transition from freelancing to social media, using content as outbound marketing and a portfolio replacement.
Tom explains how he learned by building a mobile app via an online course and emphasizes that self-teaching is common in development. The discussion covers generative AI’s productivity boost and the “gen AI wall,” arguing that non-technical builders can start fast but get stuck without domain knowledge to debug and judge code quality.
He describes his racing “pit wall” data project, where templates make AI-generated features easier, while stressing the need to review and test code.
---
Why Al Is Changing Who Gets to Build feat. Tom Shaw | AI Pulse podcast by ABBYY
00:00 Idea Is The Barrier
00:29 AI Pulse Welcomes Tom Shaw
01:10 Accidental Creator Journey
03:07 Learning To Code, Self Taught
04:25 GenAI Wall For Builders
06:34 F1 Pit Wall Templates
08:42 Trust And Review AI Code
12:34 Misconceptions And Imposter
17:07 Hype Influencers And Agents
21:28 Smoke Mirrors Open Source
23:24 Future Of Building And SaaS
27:09 Pricing Seats vs. AI Agents
28:44 Wrap Up And Last Thoughts
---Have a question you'd like us to ask our next guest? Drop it in the comments! ---
7 May 2026
Building Trust in AI, What to Look For | AI Governance Experts Arcangelo Leone de Castris and Uzma Chaudhry
Is your organization ready to navigate the evolving landscape of enterprise artificial intelligence?
In this episode of the AI Pulse Podcast, AI Governance experts Arcangelo Leone de Castris and Uzma Chaudhry define governance in two buckets: public-policy rules shaping market actors and organizational practices governing AI system behavior, emphasizing risk management, safeguards, and implementation.
Chapter Markers
00:00 Welcome and Guests
02:09 Defining AI Governance
04:58 Roadmaps and Implementation
07:06 Why Governance Matters Now
10:17 Everyday Risks and Bias
13:04 Black Boxes and Explainability
15:33 Transparency and Data Provenance
20:43 Copyright and AI Creativity
26:35 Can AI Be Creative
32:40 Star Trek and Wrap Up
---
Uzma is an AI law and policy practitioner. Her research focuses on enterprise AI governance, copyright and AI, the governance of foundation models through technical standards, and legal alignment in advanced AI systems. She has previously led research on AI governance at The Alan Turing Institute and the IAPP.
Arcangelo is the AI Data Governance Lead at AVIVA and a former Research Manager at The Alan Turing Institute, where he led the InnovateUK BridgeAI governance programme and supports the development of the AI Standards Hub. He has contributed to European standardisation for the EU AI Act within Cen-Cenelec JTC 21 and previously served as a Parliamentary Assistant at the European Parliament. Arcangelo is also a Fulbright alumnus of the University of Chicago.
This discussion traces governance momentum before and after ChatGPT, noting accelerating AI adoption, agentic AI, and the growing need for robust frameworks. They address the “black box” problem, limits of explainable AI for Large Language Models (LLMs), and offer practical advice for users, including avoiding automation bias, questioning outputs, and probing models for honest uncertainty.
If you are curious about transparency across the AI lifecycle, the EU AI Act’s disclosure expectations, and the difficulty of data provenance at internet scale – this episode is for you.
From lively dicussions on AI and creativity to copyright ambiguity, and ethical risks like cultural entropy and synthetic-data feedback loops - this episode will help you navigate the evolving tech landscape safely and effectively. Tune in to discover the essential strategies you need to stay competitive in a rapidly changing technological world.
---
Leave a comment to join the conversation. What question should we ask our next guest?
---
Helpful links
Visit the ABBYY website to explore our platform and start driving higher accuracy today. Contact us for a demo: https://www.abbyy.com/company/contact-us/?utm_source=spotify&utm_medium=social&utm_campaign=podcast-052026 (https://www.abbyy.com/company/contact-us/?utm_source=youtube&utm_medium=social&utm_campaign=podcast-052026)
Read ABBYY's latest thought leadership articles in The Intelligent Enterprise. https://www.abbyy.com/intelligent-enterprise/?utm_source=spotify&utm_medium=social&utm_campaign=podcast-052026 (https://www.abbyy.com/intelligent-enterprise/?utm_source=spotify&utm_medium=social&utm_campaign=podcast-052026 )
---
LinkedIn: Follow AI Thought Leaders
Connect with Max: https://www.linkedin.com/in/maximevermeir/
Connect with Marlene: https://www.linkedin.com/in/dr-marlene-wolfgruber/ (https://www.linkedin.com/in/dr-marlene-wolfgruber/)
Connect with Arcangelo Leone de Castris: https://www.linkedin.com/in/arcangeloleonedecastris/ (https://www.linkedin.com/in/arcangeloleonedecastris/)
Connect with Uzma Chaudhry: https://www.linkedin.com/in/u-n-c/ (https://www.linkedin.com/in/u-n-c/)
#AIPulsePodcast #PurposeBuiltAI #EnterpriseAI #AIGovernance #AlanTuringInstitute #ComputerLanguage #AgenticAI #LLMs #GenAI #ROI
Host of AI Pulse Podcast?
Claim your podcast to manage its listing and keep your show details accurate.
Pod Engine is an independent podcast discovery and analytics service and is not affiliated with or endorsed by this podcast. Artwork and show content belong to their owners. Full legal notice.
