3 things that caught our attention this week...

Winning Over the Tough Crowd

In a world where opinions are firmly rooted—politics, anyone?—persuading “rejecters” can seem impossible. This article by Richard Shotton and Will Hanmer-Lloyd, though a few years old, shares behavioral science-backed strategies that are as relevant now as ever. From understanding confirmation bias to leveraging subtle influence techniques, it’s a quick study in changing minds and bridging divides.

Three ways to convince rejecters of the value of your offer

The U.S. Car Crash Trail to Russia

In a twist on the used car market, thousands of U.S.-sourced, wrecked vehicles are winding up in Russia through a complex export chain, starting in Georgia. This BBC article by Rayhan Demytrie dives into the billion-dollar business that takes totaled American cars, repairs them in Rustavi, and re-exports them, often to former Soviet republics. Despite sanctions, a rise in car exports to countries like Kazakhstan and Armenia is fueling a route that ends back in Russia, revealing how industry creativity finds ways around geopolitical obstacles.

Thousands of cars that crash in US are ending up in Russia

Can AI Train AI? Enter the World of Synthetic Data

With real data getting harder to come by, companies are turning to synthetic data to train their AI models. Think of it as a kind of “AI-made fuel” for the next generation of machines. Companies like Meta, OpenAI, and Microsoft are already using AI-generated data to train models, and some are finding it surprisingly cost-effective. But it’s not all smooth sailing. Synthetic data has its own issues: biases in the original data carry over, and when AI starts feeding on its own creations, errors multiply, making models less reliable over time. As the industry leans on synthetic data, the question looms—are we building smarter machines or just setting the stage for an echo chamber? More on this debate to follow.

The promise and perils of synthetic data | TechCrunch