The Death of Tokenmaxxing: Why Enterprise AI is Shifting to Open Weight Models

Justin Pagano was on a phone call when he opened his investment app, and a banner appeared: we’re not calling you, and we’d never ask for this. No block, no extra step, just a nudge at the exact moment a scam would land. It’s a small thing that raises a big question for every tech leader: when does protection help people, and when does it just get in their way? Justin, Senior Director of GRC Engineering at Vanta, argues that most governance still runs on PDFs and point-in-time snapshots that can’t keep pace with cloud-scale systems or AI. We unpack why risk rarely lands with the business, how layered defences work like seatbelts and airbags, what’s lurking in your shadow AI inventory, and why fixating on risk for its own sake can quietly wreck the outcome you were protecting. Press play if your controls look fine on paper.

David grabs Will Morgan, Principal Data Scientist at the Rajasthan Royals, on the floor at Data Decoded in Manchester, the day before they share a stage. Will went from 12 years in academia, with a PhD in atmospheric science, to building a year-round analytics function for one of the founding teams of the Indian Premier League (IPL). He explains why data culture in a traditional sport has to come from the top down. That starts with an owner who has a tech background and runs through head coach Kumar Sangakkara, who makes sure the models are part of every recruitment and selection decision.

You watch an ad and wince. It’s clumsy, it’s cringe, and you can’t imagine it converting anyone. Then it racks up 300 million views and converts brilliantly. For Ada Mockute Jaime, Chief Marketing Officer at Nordcurrent, moments like that settled the argument: she is not the customer, and her taste is not the strategy. The stakes reach well beyond gaming. AI now produces ads for a couple of euros each and pushes thousands of new games out every day, so attention goes to whoever reads their audience best, not whoever has the best taste. We unpack how social listening revealed that players felt nostalgia where marketing was selling speed, how one fast-fingered player with a modest following opened up Indonesia, and why she hires for curiosity over specialist skill. If your instincts have ever overruled your data, press play.

What happens when a prediction lands on a board slide and everyone treats it as a fact? Recorded live at Data Decoded, David sits down with transformation leader and founder Lucy Goudie straight after her talk. Her argument is that truth and untruth now move at the same speed, so organisations that outsource decisions to data are really outsourcing accountability. Lucy explains why uncertainty has to travel with every prediction. She shows how a single unchallenged number can become someone’s ceiling, and why human doubt should be treated as a data point. She also uses David’s Garmin VO2 max to make a point about business: you can’t work on outcomes like time to market directly, only on the behaviours that drive them. David would love you to complete our Digital Leadership Report. To do that, just click ⁠here⁠. WhatsApp us your comments and questions on +447392087344 Tech Talks is powered by Harvey Nash, the leading global provider of technology and talent solutions. Click ⁠here⁠ to learn more about our services. The music on Tech Talks is created by London artist ⁠Lemzi⁠.

Eight years ago, Janine Luk was teaching herself to code at night and at weekends while working in marketing. The job she does now didn’t exist. Today, plenty of young people hesitate to start, convinced AI will replace software engineers before they get a foot in the door, and most leaders can’t tell them which skills will still matter. Janine, a Senior Software Engineer at Datatonic, has a blunter answer: treat technical skill as the baseline, because everyone applying for your job already has it. The edge is explaining what you built, selling it, and knowing how it makes the business money. We unpack why junior and mid-level engineers miss that, how golf got her into rooms engineers rarely reach, what male allies actually do that helps, and why she posts and walks away. Press play if you hire engineers, or you are one wondering what’s left. WhatsApp us on +447392087344! If you’re heading to Web Summit and looking for a different way to meet attendees, we’re hosting a breakfast run club on Tuesday November 10th at 7am. Come and meet some new, friendly faces: https://luma.com/mpv6bmso Tech Talks is powered by Harvey Nash, the leading global provider of technology and talent solutions. Click here to learn more about our services. The music on Tech Talks is created by London artist Lemzi.