Fewer Late Nights, Not Fewer Humans

Alastair McDermott - HumanSpark.ai

Fewer Late Nights, Not Fewer Humans is about turning AI into real productivity gains - for you and for your team. Hosted by Alastair McDermott of HumanSpark, this show is for business leaders who want to work smarter and faster with AI, and then scale those wins across their team and their organisation. It is about embedding AI into how work actually gets done, in weeks rather than quarters. Episodes cover: Where AI genuinely saves time in a working week, and where it quietly costs you time instead. How to get a team using AI well without a six-month change programme. What to do about the ethical, legal and people questions that come with it. How other leaders are making it work, in their own words. You will find interviews with operators, solo episodes on what is working right now, and short pieces on single ideas worth your time. Subscribe if you would rather have fewer late nights, rather than fewer humans!

  1. 12/05/2024

    The Dark Side of AI: Ethical Landmines Every Business Leader Needs to Know

    This is the final episode in the *Fewer Late Nights, Not Fewer Humans* archive - recorded in December 2024 - and it's a noticeably darker conversation than the rest of the back catalogue. Alastair sits down with Tom Murphy, an AI ethics and safety specialist he's known for nearly thirty years, to talk about the parts of AI adoption that rarely make it into the excited LinkedIn posts: the biases businesses inherit without noticing, the biases they create themselves, and the reputational damage that follows misuse. As Alastair warns at the top of the episode, some of it is mildly terrifying - and he isn't entirely joking. The conversation opens with the concept of "alignment" - whether an AI is actually pursuing the goal you think it is - and Tom explains, with a memorably simple Mario Brothers example, why we can never be fully certain that it is. From there the discussion moves through explainable versus black-box models, the limits of scenario testing, and why an AI's opacity is arguably no worse than a human employee's, except that AI operates at speed and scale. Tom then brings it firmly back to business reality: GDPR obligations around explaining automated decisions, protected characteristics, and how proxies quietly smuggle banned data back in. Car colour becomes a proxy for gender. Titles like Mr and Mrs become a proxy for gender. Postcode becomes a proxy for income, race, or simply a record of who was refused before. Most unsettling of all is his facial recognition example, in which a perfectly balanced, diligently curated training set still produced a system that learned to ignore an entire subgroup of people - because throwing them away improved the headline accuracy score. The second half turns practical. Tom outlines what subgroup testing looks like, why retraining always demands retesting, and why business leaders should insist on a baseline level of performance for every type of customer they serve rather than accepting an implicit "exchange rate for people." For listeners whose ambitions stop at "I just want to use ChatGPT and move faster," he explains what you inherit from a vendor's safety team, what you're still on the hook for, and why anything speaking to customers in your voice needs review. There's a strong case made for keeping a human in the loop, framed as a way to reassure anxious staff rather than replace them. Along the way: the US military's tank-recognition model that actually learned to identify blue skies, a jewellery insurance field that turned out to predict car crashes, thieves who avoid cul-de-sacs, the trolley problem tested live on 400 data scientists, a listener question on pushback from the anti-DEI crowd, an genuinely difficult dilemma about distributing an expensive cancer drug, and why Tesla hands control back to the driver at the exact moment a decision matters most. After this episode, the show picks up in the present. Find Tom at TomMurphyAI.com or on LinkedIn. Alastair's book, *An Absolute Beginner's Guide to Using AI*, is available at https://humanspark.ai/books/beginners-guide-to-ai/ The introduction to this episode is read by an AI-generated voice. The conversation itself is unedited archive audio.

  2. 11/28/2024

    Why Your AI Strategy Is Too Complicated (And How to Fix It)

    Most AI failures aren't technology failures - they're scoping failures. In this episode from the archive (November 2024), Alastair sits down with Michael Zipursky, CEO and co-founder of Consulting Success, who has advised organisations from startups to billion-dollar corporations across more than 75 industries and is the author of *The Elite Consulting Mind* and *Consulting Success*. The conversation centres on a pattern Zipursky sees repeatedly among consulting firm owners and business leaders: the instinct to build one large, ambitious AI solution spanning five parts of the business at once, rather than solving a single narrow use case first. He is candid that his own firm made these mistakes and learned from them - and argues that unless you're a well-funded technology company with genuine project management chops, the smarter path is to pick one small use case, build it, test it, learn, and only then expand. The discussion splits AI adoption into two distinct categories: custom solutions you build yourself, and the fast-growing pile of pre-existing tools you can simply buy. Zipursky recounts an unsolicited pitch from an AI SDR service that makes phone calls while pretending to be human, and uses it to raise a question many buyers skip entirely: is this AI representing your brand the way you'd represent it? He and Alastair explore how volume-driven automation can quietly degrade response rates and brand equity, why so much automated outreach may collapse under its own weight, and the strange near-future where AI talks to AI - already visible in recruitment, where applicant-side tools now fire off hundreds of personalised applications into employer-side screening systems. The pair also look further ahead at agents, AI-optimised websites and API-first interfaces, and why science-fiction ideas from *The Jetsons* to *Star Trek* have a habit of becoming product roadmaps. On the practical side, Zipursky makes a blunt case that business owners who aren't at least actively thinking about AI's implications are being reckless - not because every industry will be transformed tomorrow, but because the odds of being the next Blockbimport, Kodak or BlackBerry are non-trivial. His guidance is deliberately unglamorous: don't try to keep up with every new tool (nobody can), don't treat it as a sprint, and instead build a culture of continuous learning where every role - marketing, research, operations, leadership - is asking how AI could make their specific work better. He also offers a neat trick for the common "I don't know what my use cases are" problem: ask the AI itself, describe your role and recurring problems, request ten options, then go deeper on one. Throughout, both agree that hallucination, context, situational awareness and expert judgement mean the human expert isn't going anywhere yet - but the expert who refuses to explore might be. **Resources mentioned:** - *An Absolute Beginner's Guide to Using AI* - https://humanspark.ai/books/beginners-guide-to-ai/ - consultingsuccess.com and the Consulting Success podcast - Connect with Michael Zipursky on LinkedIn --- The introduction to this episode is read by an AI-generated voice. The conversation itself is unedited archive audio.

  3. 11/22/2024

    Why AI Adoption Starts With Your Business Model

    In this episode from the archive (originally recorded November 2024), Alastair sits down with Patrick Ward - a commercial leader with 30 years' experience, including 12 years at Microsoft, where he spent seven years with the global IoT team and reshaped the company's approach to IoT and AI engagements. Now founder of Iteria Partners and a part-time lecturer at UCD Smurfit Business School, Patrick makes a case that Alastair has repeated ever since: successful AI implementation starts with the business model, not the technology. Get that order wrong, and no amount of tooling will save the project. Patrick explains why so many IoT initiatives at Microsoft died after a successful proof of concept. The technology worked - but the moment an organisation connected its MRI scanners, blast furnaces or HVAC systems to the cloud, its value proposition, revenue model, sales incentives, channel strategy and in-house capability all had to change. Nobody had agreed to that upfront, so the project quietly got parked behind seventeen other priorities. His answer was to run a business model workshop at the outset of every customer engagement, getting the custodians of the business model - product, sales, marketing and finance - into one room. The most commonly cited benefit? Alignment. Ask ten leaders to define the value proposition and you'll routinely get ten different answers. He shares a live example of a global HVAC business struggling to get its sales team to attach £5k-per-month digital services to a multi-hundred-thousand-dollar hardware sale, and explains why bolting a digital business onto the side of an existing one rarely works. The conversation also covers Patrick's research interviewing heads of AI at 15 multinationals operating in Ireland, where the surprising finding was that scarce data science talent wasn't the biggest constraint - educating business leaders to think about AI in the context of strategy was. He walks through the AI envisioning sessions he runs for SMEs and enterprises, from competitor analysis via job postings to facilitated brainstorming on horizontal and industry-specific use cases. On productivity, his advice is blunt: don't build a business case you can't measure, roll out piecemeal rather than enterprise-wide, and invest in governance, policy, training and sharing. He closes with a memorable story about surveying a software team's real AI usage - twenty distinct use cases in a single morning, from generating synthetic French customer data to translating a Brazilian developer's written English - plus a client whose intermittent, months-long software bug turned out to be the number one finding in a ChatGPT code review. As Patrick puts it: yes, there's hype, but there's also a lot of real value, and the job is knowing the difference. **Guest:** Patrick Ward, Founder, Iteria Partners - patrick@iteriapartners.com, or find him on LinkedIn. **Mentioned:** Alastair's book, *An Absolute Beginner's Guide to Using AI* - https://humanspark.ai/books/beginners-guide-to-ai/ --- The introduction to this episode is read by an AI-generated voice. The conversation itself is unedited archive audio.

  4. 11/19/2024

    Why This Business Owner Says AI Will Change Everything

    In this episode from the archive (originally recorded November 2024), the host speaks with Ethan Wadsworth, Director of Sales and Marketing at Discrete Heat - the family manufacturing business behind ThermoSkirt, the heated skirting board that appeared on Dragons' Den. With a team of around 20-25 people and a £3m turnover, Discrete Heat cut customer response times by 75% while *improving* service quality. No pilots, no proofs of concept, no consultants - just a small company shipping. Ethan explains how his father invented the product, how he joined at 18 and did everything from packing boxes to installations, and why demand has exploded alongside the UK's push toward net zero and heat pump adoption. The conversation gets unusually concrete about implementation. Ethan walks through how Discrete Heat's in-house developer built a bespoke CRM that calls the OpenAI API to read previous email correspondence and draft replies - turning static templates into what he calls "smart templates" that a human then approves and sends. He describes his personal workflow of talking to ChatGPT's advanced voice mode hands-free in the car, dumping unstructured thoughts, then asking it to condense them into a blog post in his own tone of voice, backed by a custom GPT loaded with samples of his actual writing. He also shares how the company hit 100,000 Instagram followers and four to five million views a month by convincing installers that rough, authentic iPhone footage beats polished content - outperforming competitors with marketing budgets in the millions. Ethan is blunt about the bigger picture: anyone calling AI a fad is "about to be swept off the pier by a massive tidal wave." His motorway analogy - AI as a fourth lane opening up while big businesses sit in traffic asking whether it's allowed - captures why he thinks small firms have the advantage right now. The discussion also covers the human side: how removing the "dreary job nobody wants" from customer service created genuine buy-in, with staff now coming to him asking for new AI features rather than resisting them. The host shares a parallel case study from a blood testing lab where data entry dropped by 90%, freeing highly trained scientists to do actual analysis. They close on education, the risk of false plagiarism accusations, an economy built on the scarcity of knowledge, and Ethan's one piece of advice for business leaders: don't search YouTube for "how to use AI" - start with your pain points, fix one, and let the penny drop for everyone else. Discrete Heat can be found at DiscreteHeat.com. The introduction to this episode is read by an AI-generated voice. The conversation itself is unedited archive audio.

  5. 10/30/2024

    The Smart Way to Automate: How to Implement AI That Actually Works For Your Business with Heidi Araya

    Are you drowning in AI hype but still waiting to see real, tangible results for your business? You're not alone. In this episode of Fewer Late Nights, Not Fewer Humans, host Alastair McDermott sits down with Heidi Araya, a process improvement consultant with over 25 years of experience driving multi-million dollar improvements through AI automation. This insightful conversation cuts through the noise to reveal how businesses are using AI in the real world to save time, increase profitability, and gain competitive advantage. Through practical examples and case studies, Heidi shares her expertise on finding genuine value in AI implementation. Key topics include: Why many businesses struggle to find value in AI implementationsHow to identify the right AI use cases for your businessReal-world examples of successful AI automationThe future of AI in sales and customer servicePractical approaches to AI implementation without eliminating jobs Tune in to discover how you can move beyond the hype and start implementing AI solutions that deliver real results for your business. Key Insights Many businesses struggle with AI because they focus on the tool first instead of identifying their specific business challengesRecording meetings and calls can provide valuable data for AI processing, though privacy considerations need to be addressedAI automation should focus on enhancing human capabilities rather than replacing workersSmaller businesses may be better positioned to implement AI while maintaining jobs, as they need diverse skill setsThe capabilities of AI are rapidly evolving, requiring regular reassessment and experimentation Strategies Start by identifying pain points: "What makes you groan when you wake up?"Use AI for automating repetitive tasks while maintaining human touch for important interactionsImplement proof of concepts before making large investments in AI solutionsConsider privacy implications when selecting AI tools and platformsRecord and analyze sales calls to improve performance and understanding of client needs Topics include: AI implementation, process improvement, business automation, sales automation, customer service, ai strategy, small business, practical ai applications, business transformation Speakers: Heidi Araya Host: Alastair McDermott Quotes "So first of all, I ask them, what makes you groan? You wake up in the morning, you're ready to start your day, what makes you groan?" - Heidi Araya, 3:08, Fewer Late Nights, Not Fewer Humans "It may not be today, it may not be tomorrow or even next year, but three years from now, they will wish that they had taken more action and kept on the forefront of it." - Heidi Araya, 29:33, Fewer Late Nights, Not Fewer Humans Fewer Late Nights, Not Fewer Humans 🎙️+📺 SHOW: Fewer Late Nights, Not Fewer Humans is about turning AI into real productivity gains - for you and for your team. Hosted by Alastair McDermott of HumanSpark, it is for business leaders who want to work smarter and faster with AI, and then scale those wins across their team and their organisation. Join me to learn how AI can help you automate repetitive tasks, boost productivity, and create a more enjoyable workplace for your team. 📲 | FOLLOW on YouTube: Fewer Late Nights, Not Fewer Humans  🎓 COACHING: Explore opportunities to work with me: AI Thought Leadership Coaching 👑 BOOKS: Discover more by searching "Alastair McDermott" on Amazon 💻 - WEBSITE: https://humanspark.ai

  6. 09/26/2024

    AI and the Future of Thought Leadership

    This episode is a bit different. What you’ll hear is a demo of Google Notebook LM. I gave it the entire 40,000 word manuscript of my book - at the time called The AI-Powered Thought Leader, now titled Use AI. Stay Human. The tool then created this AI-generated conversation between two hosts, based entirely on the book text. That’s what you’ll hear in this episode - the raw output from NotebookLM. It’s an example of what these new AI tools can do when you hand them a big chunk of text. What to expect Two AI hosts talking through the ideas in my book.No editing, explanation, or polish - this is straight from Notebook LM.A glimpse at how AI can generate summaries and overviews from long-form content. Why I shared this I wanted to be transparent: this is AI-only audio. It’s not perfect, but I thought it would be interesting to share what Notebook LM produced when I fed it a full book draft. If nothing else, it gives you a sneak peek at some of the content of my book Use AI, Stay Human. Quotes "AI can write you a novel, but it can't write your novel." "It's not about becoming a robot, but about becoming even more human." - Guest speaker This emphasises the importance of enhancing our human qualities as we integrate AI into our work. Useful Resources "Use AI, Stay Human" by Alastair McDermott - A comprehensive guide to using AI in thought leadershipChatGPT and other AI writing tools - Practical AI assistants for content creationAI tools for research and data analysis - Resources to help you process information more efficiently About Me I'm Alastair McDermott. I work as a podcaster, writer, and consultant, focusing on the use of AI in thought leadership. My book, "Use AI, Stay Human," offers a practical roadmap for using AI to enhance your expertise and increase your influence in the digital space. I've spent years exploring the intersection of technology and thought leadership, and I'm excited to share these insights with you. Next Steps If you're keen to power up your thought leadership with AI, you can read "Use AI, Stay Human" at humanspark.ai/books/use-ai-stay-human. It's packed with actionable strategies and real-world examples. I also invite you to join the conversation on social media using #AIPoweredThoughtLeader. Share your thoughts, experiences, and questions about AI in thought leadership.

  7. 08/09/2024

    AI for SMEs: How to Thrive in the New Business Landscape

    Is your business ready for the AI revolution, or are you still stuck in the dark ages? In this eye-opening episode of Fewer Late Nights, Not Fewer Humans, host Alastair McDermott sits down with AI implementation expert Eric Bye to examine how businesses of all sizes can use AI. Eric shares his wealth of experience in helping companies navigate the complex landscape of AI integration, offering practical advice on how to get started, avoid common pitfalls, and maximize the benefits of this fantastic technology. From identifying key workflows to addressing data privacy concerns, this episode is packed with actionable insights for business leaders looking to stay ahead of the curve. Some of the valuable topics discussed include: How to identify AI opportunities within your existing business processesThe importance of benchmarking and quantifying AI improvementsStrategies for overcoming employee resistance to AI implementationThe future of AI in business, including voice interaction and automation Whether you're an AI skeptic or an early adopter, this episode will challenge your assumptions and provide a clear roadmap for using artificial intelligence in your business. Tune in now to ensure you're not left behind in the AI revolution! Topic: ai implementation, business processes, workflow optimization, data privacy, employee training, ai adoption, technology integration, business efficiency, automation, digital transformation Speakers: Guest: Eric Bye. Host: Alastair McDermott. Key Insights: AI implementation should focus on augmenting and transforming jobs rather than replacing themBusinesses should start by mapping out their workflows and identifying areas where AI can add valueData privacy and security are crucial considerations when implementing AI solutionsMany businesses are informally using AI through personal accounts, which poses risksAI can improve both efficiency and the quality of outputs in various business processesThe future of AI in business includes more automation and proactive task managementExperimenting with AI tools is key to understanding their capabilities and limitations Arguments: AI implementation requires a collaborative approach to overcome employee skepticism and resistanceEnterprise-level AI solutions may be too complex and expensive for many SMEsThe current state of AI agents is still too clunky for widespread business adoptionWaiting for "perfect" AI solutions may cause businesses to fall behind competitors Conclusions: Businesses should start experimenting with AI now to build a foundation for future advancementsA balance must be struck between using AI capabilities and maintaining data privacyAI implementation should be tailored to each business's specific needs and risk profileOngoing education and training are essential for successful AI adoption in businesses Quotes: Eric Bye, [17:29]: "If you ask an employee to create a presentation for, say, a pitch, there's a lot of there's something there's hundreds of 1000s of words or tokens that are unsaid, when you ask them to create that they understand your relationship with the client, they understand the history, they're probably see seed on a whole bunch of emails, they understand what's worked, and what hasn't in the past." - Fewer Late Nights, Not Fewer Humans Eric Bye, [36:45]: "There's tons of opportunity today. And I think some people would like to wait until they they see it as perfect. But it's just different technology that operates differently, which makes people feel a little bit uncomfortable." - Fewer Late Nights, Not Fewer Humans Find Eric at https://erictronai.com Find Alastair at

  8. 08/08/2024

    How to Actually Use AI to Drive Business Growth

    Everyone is talking about AI, but plenty of businesses still can't point to a return they'd be willing to defend in front of whoever controls the budget. In this episode from the archive (recorded August 2024), Alastair sits down with Alvaro Melendez, founder of Krant Creativity and Technology, to tackle the least glamorous question in the field: where is the money? Not the demo, not the pilot - the measurable result. They use the widely-reported Chevron example as a starting point: tens of thousands of licences, $30–$40 per user per month, and no visible ROI. Alvaro's diagnosis is blunt - if a $20-a-month assistant can't save you $20 a month, the use case is wrong, and that's usually not the employee's fault. The conversation digs into why so many organisations end up with expensive experiments instead of results. Alvaro argues the core mistake is thinking of AI as a tool rather than as a thinking partner - what he calls a *dupla*, borrowing from the creative-agency pairing of copywriter and art director. Asking a frontier model to shave ten percent off email reading time is, in his framing, hiring Albert Einstein and asking him to serve coffee. He offers a sharper reframe for leadership: you didn't buy 20,000 licences, you hired 20,000 very smart employees - so what are they actually going to do? From there the two work through practical territory: breaking jobs down into tasks to see which ones AI genuinely improves, using better creative output to win pitches that would otherwise have been lost, and running hands-on exercises (his Titanic dataset workshop for senior leaders is a standout) that permanently change how executives think about the technology. They also cover the guardrails. Alastair walks through the failure modes people only learn by doing - context windows degrading, and the harder-to-spot problem of plausible fabrications, including the New York lawyer whose invented case citations got him thrown out of court. Alvaro makes the case for an internal AI council rather than outsourced expertise, for bringing ethics and legal in from day one, and for starting with the ten people who actually want to be there instead of trying to convert all hundred. Along the way: the Gemini Olympics ad that demonstrated exactly the wrong use case, a father-in-law learning to code in R over a weekend, real-time Irish translation at a workshop on the west coast of Ireland, and a closing list of the things Alvaro refuses to hand over to a machine - the letter to Santa, the birthday cake, the bedtime story. *Two notes on the closing minutes: Alastair's call for beta readers closed long ago, and the book he names as "The AI-Powered Thought Leader" was published as **Use AI, Stay Human**.* The introduction to this episode is read by an AI-generated voice. The conversation itself is unedited archive audio.

See All (24)

About

Fewer Late Nights, Not Fewer Humans is about turning AI into real productivity gains - for you and for your team. Hosted by Alastair McDermott of HumanSpark, this show is for business leaders who want to work smarter and faster with AI, and then scale those wins across their team and their organisation. It is about embedding AI into how work actually gets done, in weeks rather than quarters. Episodes cover: Where AI genuinely saves time in a working week, and where it quietly costs you time instead. How to get a team using AI well without a six-month change programme. What to do about the ethical, legal and people questions that come with it. How other leaders are making it work, in their own words. You will find interviews with operators, solo episodes on what is working right now, and short pieces on single ideas worth your time. Subscribe if you would rather have fewer late nights, rather than fewer humans!

Information

  • Creator
    Alastair McDermott - HumanSpark.ai
  • Years Active
    2024
  • Episodes
    24
  • Rating
    Clean
  • Copyright
    © Copyright 2023-2026 HumanSpark AI Ltd. All Rights Reserved.
  • Show Website
    Fewer Late Nights, Not Fewer Humans