Welcome to Episode 2 of Inventing Fire With AI. This episode is taken from a discussion I had at Medellin Tech Talk ‘25 in Colombia with Jorge Armando González Pino, Camilo Jaramillo, and Luis Robles. AI isn’t just a question of which company will emerge as the winner, but which country.

AI is replacing search engines, but also, people are telling their AI chatbots their most personal information, from medical information to their deepest fantasies. And yet, we have little to no guardrails against this information being stolen or exploited.

On jobs, disruption won’t hit everyone equally. Expect significant role shifts and net new opportunities—if you’re learning AI now. Some hedge with “blue-collar moats” (e.g., plumbing/electrical businesses), while others double down on AI mastery.

The near future isn’t only cloud. With AI chips in phones and laptops, smaller on-device models and “AI of Things” will grow, enabling private, low-latency use cases. But AI’s “kryptonite” is real: energy and water demands for data centers, infrastructure fragility (pull the plug), and compounding data-quality issues as models train on AI-generated content.

Bottom line: mix your tools, iterate fast, validate results, and build AI literacy now—because the winners will be those who put in the work.

#AI #Mill5 #InventingFire #Claude #ChatGPT #GitHubCopilot #Gemini #Kagi #LLM #AIEthics #Privacy #Data #OnDeviceAI #OpenSource #PromptEngineering #Coding #Productivity #Jobs #DataCenters #Energy #Water