How People Really Use ChatGPT | Lessons from Zuckerberg’s Meta Flop | MIT’s Research on AI Romance

Happy Friday Everyone! I hope you've had a great week and are ready for the weekend. This Weekly Update I'm taking a deeper dive into three big stories shaping how we use, lead, and live with AI: what OpenAI’s new usage data really says about us (hint: the biggest risk isn’t what you think), why Zuckerberg’s Meta Connect flopped and what leaders should learn from it, and new MIT research on the explosive rise of AI romance and why it’s more dangerous than the headlines suggest.If this episode sparks a thought, share it with someone who needs clarity. Leave a rating, drop a comment with your take, and follow for future updates that cut through the noise. And if you’d take me out for a coffee to say thanks, you can do that here: https://www.buymeacoffee.com/christopherlindWith that, let’s get into it.⸻The ChatGPT Usage Report: What We’re Missing in the DataA new OpenAI/NBER study shows how people actually use ChatGPT. Most are asking it to give answers or do tasks while the critical middle step, real human thinking, is nearly absent. This isn’t just trivia; it’s a warning. Without that layer, we risk building dependence, scaling bad habits, and mistaking speed for effectiveness. For leaders, the question isn’t “are people using AI?” It’s “are they using it well?”⸻Meta Connect’s Live-Demo Flop and What It RevealsMark Zuckerberg tried to stage Apple-style magic at Meta Connect, but the AI demos sputtered live on stage. Beyond the cringe, it exposed a bigger issue: Meta’s fixation on plastering AI glasses on our faces at all times, despite the market clearly signaling tech fatigue. Leaders can take two lessons: never overestimate product readiness when the stakes are high, and beware of chasing your own vision so hard that you miss what your customers actually want.⸻MIT’s AI Romance Report: When Companionship Turns RiskyMIT researchers found nearly 1 in 5 people in their study had engaged with AI in romantic ways, often unintentionally. While short-term “benefits” seem real, the risks are staggering: fractured families, grief from model updates, and deeper dependency on machines over people. The stigmatization only makes it worse. The better answer isn’t shame; it’s building stronger human communities so people don’t need AI to fill the void.⸻Show Notes:In this Weekly Update, Christopher Lind breaks down OpenAI’s new usage data, highlights the leadership lessons from Meta Connect’s failed demos, and explores why MIT’s AI romance research is a bigger warning than most realize.Timestamps:00:00 – Introduction and Welcome01:20 – Episode Rundown + CTA02:35 – ChatGPT Usage Report: What We’re Missing in the Data20:51 – Meta Connect’s Live-Demo Flop and What It Reveals38:07 – MIT’s AI Romance Report: When Companionship Turns Risky51:49 – Final Takeaways#AItransformation #FutureOfWork #DigitalLeadership #AIadoption #HumanCenteredAI