Two Millennials and Mom
What happens when a tool does exactly what you asked…just not what you meant? This week, Callie, Cole, and Mecca wade into the murky, fast-moving waters of artificial intelligence: not the robot apocalypse version, but the version that's already here, already making decisions, and already raising questions none of us have fully answered. From the classic “paperclip problem” to real-world legal debates and an AI-run retail store that’s already lost $13,000, we explore the tension between innovation and risk. The conversation stretches from philosophical to deeply practical. Who's responsible when AI assists in something harmful? Can we hold a corporation accountable the same way we hold a person? And perhaps most urgently, does anyone actually have their hand on the wheel? This is a conversation about curiosity, unease, and the very human tension between progress and caution. 10,000-Foot View of this Episode: * The Paperclip Problem: The hosts unpack a famous thought experiment…an AI told to make paperclips that eventually consumes everything in pursuit of its single goal. Callie finds it chilling; Cole calls it oversimplified; Mecca wonders if any of us are smart enough to set the right parameters in the first place. * The Genie Problem: Cole reframes AI alignment as negotiating with a folklore genie; one that will grant your wish exactly as asked, consequences be damned. It's a useful lens for understanding why good intentions and bad outcomes aren't mutually exclusive. * The Arms Race No One's Winning: The conversation draws a sharp parallel between AI development and the nuclear race except this time there are thousands of players, not a handful of nations, and the economic stakes mean everyone is incentivized to move faster, not slower. * Who Gets Sued When AI Helps Plan a Crime? A live court case involving ChatGPT and a school shooter puts the liability question front and center. The hosts wrestle with where the line falls…between tool and weapon, between prompt and intent, between a gun manufacturer and the person who pulled the trigger. * The Luna Experiment: Callie's weird thought delivers a real-world case study: Luna, an AI given $100,000 and a mandate to run a San Francisco retail store. She bought a lot of candles. She forgot to schedule her employees. She's $13,000 in the hole. And she's powered by Anthropic's Claude. The hosts can't decide if it's a disaster or just a startup. * The Entry-Level Cliff: If AI takes over entry-level jobs, where do young workers learn? Cole thinks they'll have to find something else to do. Mecca thinks it's more nuanced than that. Callie thinks we should probably have that conversation before we need it. * Paying Attention Is the Bare Minimum: All three hosts land in the same general place. Not panic, not naïve optimism, but a commitment to staying informed, asking harder questions, and holding both themselves and AI developers accountable for how this technology gets used. Memorable Quotes: * "If we can get one new product that's going to last for two or three years and the risk is existential crisis, I don't think that's worth it." – Cole * “There are no guardrails in place at all yet. This is all dependent upon the good humor or the values of whoever is creating this.” – Mecca * “I kind of see AI like a teenager that's going to do whatever it wants to do, and it's going to do whatever it thinks that it needs to do to sneak one past you, to pull the wool over your eyes.” – Callie * "From a corporate standpoint, they know how powerful [AI] has the potential to be. And if they're not the first ones to it, that means they lose. That could mean that their corporation gets gobbled up by whoever gets there first." – Cole * “Even the experts don't agree on where we're going with [AI] and what the dangers are.” – Mecca * “At what point are we going to sit back and say, 'guys, the potential to burn the house down is really, really prevalent here. We probably shouldn't do this.'?” – Callie * "Is the risk of failure or adversity worth not trying?" – Cole * “[AI is] a really big deal and we need to be at least paying attention.” – Mecca * “When is a corporation a person versus not a person? Well, they're not a person when it's convenient for them to be prosecuted, but they are a person when it's convenient for us to get our way with spending money on political action campaigns…” – Callie Resources Mentioned: * Read up on the paperclip maximizer [https://tmampod.short.gy/zWUANW] thought experiment (originally from philosopher Nick Bostrom) * Luna is the AI-run retail store in San Francisco [https://tmampod.short.gy/xBKfXV] that's powered by Anthropic's Claude. It's currently the subject of a three-year experiment in autonomous business management. * The ChatGPT school shooting lawsuit [https://tmampod.short.gy/TVmgXj] is an active legal case examining whether OpenAI bears liability for responses that may have assisted in planning a mass shooting. * We're big fans of repealing Citizens United v. FEC. [https://tmampod.short.gy/PVHrVF] It's the Supreme Court case that granted corporations First Amendment speech rights, which Callie argues creates a confusing double standard when it comes to corporate accountability. We're all navigating this in real time. That includes the people building it. So the question this week isn't whether you trust AI. It's whether you're paying attention to how it's being used, who's using it, and what we owe each other in the meantime. Start there. And if you've got thoughts on where the line is or whether we've already crossed it, we want to hear from you. If this episode got you thinking, share it with someone who’s trying to make sense of AI too. And don’t forget to follow Two Millennials and Mom wherever you get your podcasts so you never miss a conversation.
83 episodios
Comentarios
0Sé la primera persona en comentar
¡Regístrate ahora y forma parte de la comunidad de Two Millennials and Mom!