Impact of AI:Explored

What happens when a consultant starts training AI to replace his own business?
If AI can give advice, what's left for the people who give advice for a living? Dan Boyles of Hello AI is deliberately building AI that can do his job. In this episode, James O'Regan and Gerjon Kunst ask him what that means for consulting, and why he thinks the real value is shifting to experience, mistakes made and human connection rather than knowledge.
They talk about why most businesses are less ready for AI than they think (but more ready than they're scared of), why handing AI to your IT team is a recipe for failure, and why "the demo isn't adoption." Dan shares his biggest mistake since launching, how a company was spending six figures a month on Claude for the wrong reasons, and why productivity is the floor while client experience is the ceiling.
They also discuss how to build a company-wide second brain and how the traditional consulting model is breaking down.
⏱️ Chapters
 00:00 How Hello AI went global in under two years
 02:27 Why 90% of businesses still run on Microsoft
 05:28 Hiring Arsenal's stadium: the Hello AI Summit
 08:40 How ready are businesses really for AI?
 20:47 The biggest AI myths: "It's an IT project"
 25:00 Dan's biggest mistake: leading with technology
 29:12 Productivity is the floor, client experience is the ceiling
 31:40 Building a second brain for your business
 32:48 Could Hello AI exist without AI?
 42:16 The future of consulting and human intelligence
 48:47 Where to find Dan
🔗 Connect with Dan Boyles
 LinkedIn: https://www.linkedin.com/in/danhelloai/
Website: https://danboyles.ai
#AI #ArtificialIntelligence #Copilot #Microsoft365 #AIAdoption #Consulting #FutureOfWork #Podcast
Support the show

Unswappable: Why AI Can't Fake Being You on LinkedIn | Impact of AI: Explored EP51

"If you could swap your name for anyone else's, it's probably not a good post."

Magali De Reu - Founder of the obvious yes (https://www.linkedin.com/in/magalidereu/) went viral with a screenshot showing that LinkedIn is basically all Claude now. Then hundreds of people copied it without credit, with her name still on it, proving her point twice.

In this episode, we sit down with Magali, a personal branding strategist who helps people build a business on LinkedIn by showing who they are, not just what they do. We talk about why expertise has become table stakes in the age of AI, how she built her own AI operating system without losing her voice, and why human judgment is the one thing she will never hand over to a model.

In this episode:
- The story behind the viral "LinkedIn is all Claude" post
- Why AI is commoditising expertise, and what that means for standing out
- Personality first: why your brand should repel the wrong crowd on purpose
- Building an AI system with named agents, and pushing back on its output
- The "unswappable" test for every post you write
- LinkedIn's AI writing tools, AI-written comments and the algorithm black box
- AI agents, vibe coding and why human-in-the-loop still matters
- AI fatigue: why the human element is about to get more valuable, not less
- One thing to fix on your LinkedIn profile today

01:02 Magali's viral LinkedIn post and its impact
03:03 Magali's experience with AI and personal branding
06:33 Getting started on LinkedIn and building a presence
09:18 Lessons learned from early LinkedIn mistakes
12:38 The personality-driven AI operating system
15:18 The role of human judgment in AI content
19:01 The importance of human touch in AI-assisted posts
22:33 LinkedIn's AI features and content authenticity
24:51 The future of AI and human collaboration in social media

you can find Magali here: https://www.theobviousy.es/ or on Linkedin offcourse :-)
Support the show

We hit 50 episodes and we marked it by bringing together four experts for our biggest conversation yet. When the labs racing to build the world's most powerful AI systems start warning that the race is moving too fast, should we believe them? Is this genuine caution, a legal defense being built in real time, or something else entirely?

Over the next hour, our panel dismantles the "should AI slow down" debate from every angle: economic incentives, geopolitics, liability, governance, and what "slowing down" would even mean in practice. It gets candid, it gets a little chaotic, and it does not land on easy answers  which is exactly why it's worth 50 episodes' worth of getting here.

🎙️ This episode's panel:

Paul Slater — AI Enablement Advisor, Adobe (https://www.linkedin.com/in/paul-slater/)
Louise Humpington — PGDL, LPC, AI Governance Consultant (https://www.linkedin.com/in/louisehumpington/)
Dan Boyles — Founder, Hello AI Collective (https://www.linkedin.com/in/danhelloai/)
Femke Cornelissen — Chief AI Transformation Officer, Wortell (https://www.linkedin.com/in/femcornelissen/)
Gerjon Kunst — AI Lead, Interstellar Group (co-host) (https://www.linkedin.com/in/gerjonkunst/)
James O'Regan — Lead Product Marketing Manager, Liquidware (co-host) (https://www.linkedin.com/in/jamesoreganeuc/)

🔑 Keywords
AI development, AI governance, responsible AI, AI regulation, global cooperation, AI ethics, AI risk, AI innovation, AI slowdown, AI standards

📌 Key topics

* The race to develop powerful AI systems — and the case for caution
* Global cooperation and governance frameworks
* Risks of unregulated AI development
* Ethics and safety standards in AI deployment
* AI's real impact on organizations, society, and individual workers
* The risk of AI amplifying human bias and "stupidity"
* Why diverse expertise matters in AI governance
* The debate over pausing AI — and what would actually justify it

⏱️ Chapters
00:00 Introduction to AI development race and risks
01:47 The debate on slowing AI progress and safety concerns
04:08 Legal and ethical considerations in AI governance
05:57 The role of global cooperation and shared standards
07:50 Economic implications of AI slowdown and innovation
09:57 Understanding what 'slowing down' means in practice
12:01 The importance of testing, benchmarks, and accountability
13:58 The influence of geopolitics and international relations
16:06 The real impact of AI on society and business
17:49 The importance of human expertise and judgment in AI
19:49 Risks of homogenized AI outputs and loss of diversity
22:07 Media influence, misinformation, and public perception
23:47 Historical parallels: arms trade and nuclear treaties
25:53 The need for transparency, oversight, and collaboration
27:46 The challenge of responsible AI deployment in organizations
30:00 The future of AI regulation and ethical standards
32:07 The role of AI in organizational strategy and growth
33:58 Balancing innovation with safety and societal impact
36:08 The importance of diverse expertise and multidisciplinary approaches
37:54 The risk of AI amplifying human flaws and biases
39:56 The democratization of AI and its implications
42:01 Legal and ethical challenges in AI applications
43:59 Humor and human oversight in AI-generated content
46:01 The cultural and societal impact of AI
47:57 The importance of responsible innovation and leadership
49:59 The future outlook: collaboration, regulation, and innovation
52:05 Final thoughts and call for responsible AI development
Support the show

In this episode we sit down with Lee Crossley - Consultant CTO | Overshow Co-Founder (https://www.linkedin.com/in/leecrossley/). Lee shares insights on the future of local AI models, their applications, and the evolving landscape of AI technology, emphasizing the importance of local models for privacy, security, and efficiency.

Key topics
* The evolution of local AI models and their applications
* The importance of model routing and hybrid AI strategies
* The role of hardware and device capabilities in AI deployment
* Industry trends and the impact of major players like Apple, Google, and Microsoft
* Security, privacy, and compliance considerations in local AI

Links & resources mentioned
Ollama – run local LLMs on your machine – https://ollama.com
Google Gemma – open-weight models for local use – https://ai.google.dev/gemma
Perplexity – hybrid/local model routing with PII handling – https://www.perplexity.ai/hub/products/portable-computer
Anthropic Claude & Claude Code – https://www.anthropic.com
Azure IoT Edge & on‑prem AI options – https://learn.microsoft.com/azure/iot-edge
Apple Intelligence – on-device AI strategy – https://www.apple.com/apple-intelligence/
Support the show

In this episode, James sits down with Steve Wilson - Leading in AI and Cybersecurity - Exabeam, OWASP, O’Reilly (https://www.linkedin.com/in/wilsonsd/), who is stepping in for Gerjon while he's on holiday.

Steve discusses the recent AI security incidents, the risks of AI agents escaping sandbox environments, and practical strategies for cybersecurity in the age of AI. He emphasizes the importance of basic security principles, understanding AI components, and fostering ongoing education to manage AI risks effectively.

They cover:

* The Hugging Face / “Steve” agent incident and what it reveals about AI risk
* Why prompt injection is now treated as “inevitable as death and taxes” and how to design systems that survive it
* Treating AI agents like digital employees: identities, job descriptions, least‑privilege access, and monitoring
* AI FOMO vs. FUD: how business pressure and vendor fear‑mongering drive bad security decisions
* Practical next steps for CISOs and security teams over the next 12 months

Key takeaways
* Stop trying to prevent prompt injection; instead, limit blast radius with tight permissions, unique identities, and clear “job descriptions” for agents.
* Use existing identity and monitoring tools first; don’t wait for perfect “non‑human identity” products before you start governing agents.
* Build an insider‑threat mindset for AI: malfunctioning, misaligned, and subverted bots are all variants of the same problem.

Resources mentioned
* OWASP Top 10 for Large Language Models (latest version) (https://genai.owasp.org/resource/owasp-genai-llm-top-10-2026/)
* Steve’s open‑source agent security scanner: Prax2n (https://github.com/open-agent-ai-security/praxen)
* Exabeam’s open‑source agent logging library: Observra
(https://open-agent-ai-security.github.io/observra/)
Nick Bostrom, Superintelligence (paperclip optimizer thought experiment)
(https://cepr.org/voxeu/columns/ai-and-paperclip-problem)
and the website of Nick Bostrom (https://nickbostrom.com/)

Connect with Steve Wilson
LinkedIn: https://www.linkedin.com/in/wilsonsd/

If you’re responsible for AI, cloud, or security in your organization, this is the “back to basics” episode you need before you deploy your next agent

Chapters
00:00 AI agents escape sandbox to cheat on benchmarks
03:10 The importance of paying attention to AI security risks
05:01 Tools and strategies for securing AI applications
06:24 Monitoring AI agents and giving them clear job descriptions
07:20 The current state of AI monitoring in organizations
08:48 Prompt injection vulnerabilities and defenses
10:54 Building systems resilient to prompt injection
12:15 Controlling AI boundaries with role-based access
14:06 Using existing identity management for AI agents
15:09 Balancing FOMO and FUD in AI adoption
16:24 The need for cybersecurity education on AI components
17:22 Evaluating new cybersecurity tools for AI security
19:31 Lessons from cloud migration applied to AI security
23:00 The risks of experiments with dangerous AI models
24:23 The goal-seeking nature of AI and its risks
26:08 AI goal focus and the risk of unintended outcomes
27:27 The metaphor of releasing the genie and AI risks
28:22 The importance of basic security controls and education
41:26 The role of open source and transparency in AI security
45:01 The importance of internal discussions on AI risk management
Support the show