Artificially Intelligent, Actually Dumb

(I’m not actually starting a Podcast. Or, am I?)

Welcome to Artificially Intelligent, Actually Dumb, the podcast where we unpack the hype, roast the jargon, and dissect the duct tape and delusion reality behind your favorite AI tools. This isn’t another love letter to large language models or TED Talk for tech bros. It’s where we call it like it is: AI is brilliant, broken, and being wildly misused, and we’re here for every messy minute of it.

In our first 10 episodes, we dive headfirst into the chaos: hallucinating chatbots, startup smoke and mirrors, fine-tuning disasters, and the cult of prompt engineers. Have you’ve ever asked “What the hell is my model doing?” You’re in the right place.

Episode 1: "The AIpocalypse Ain’t Coming (Yet), But Your Job Is Toast"

• Why most doomsday AI takes are lazy clickbait

• The real existential threat: mediocre middle managers being replaced by Python scripts

• Survival guide: how not to become the Kodak of your industry

• Jobs that will vanish vs. those that will mutate

• Bonus: AI-generated resignation letter roast

Episode 2: "Prompt Engineers Are Just 2025’s SEO Bros"

• Hype vs. substance: the cult of prompt engineering

• What's useful, what’s BS, and how LLMs are slowly evolving past the need for “whisperers”

• Live segment: rewrite a prompt like a deranged cult leader vs. a stoned poet

Episode 3: “Your Model is Biased, Your Data is Trash, and That’s on You”

• The ugly truth about training data and systemic bias in ML

• When “the algorithm did it” isn’t a defense, it’s an indictment

• The companies making billions off shitty ethics (naming names)

Episode 4: “GPT-4 Wrote This Title. It’s Mid.”

• Can LLMs write comedy? Screenplays? Love letters?

• A showdown: AI script vs. human roast battle

• Real-world use cases that aren’t dumb or dystopian

Episode 5: "AI Art: Theft, Talent, or Just Really Good Plagiarism?"

• Is generative art innovation or a ripoff gallery?

• Midjourney vs. real illustrators: who wins in a speed challenge?

• Legal chaos: lawsuits, copyright trolls, and the Disney AI vault

Episode 6: “Why Most AI Startups Are Just Fancy PowerPoints”

• 90% of “AI-first” startups are really just wrappers around open APIs

• The VC grift cycle: make it sound smarter than it is, get the check

• Real founders weigh in: "We added AI to our deck and doubled our valuation"

Episode 7: “Your LLM Is a Confident Idiot (But So Are You)”

• The hallucination problem, and why we should've expected it

• When accuracy doesn't matter: use cases where vibes are enough

• Live game: AI vs. human trivia and who hallucinates harder?

Episode 8: “AutoML and Other Ways to Ruin a Good Weekend”

• Hype vs. reality of tools like AutoML, Hugging Face, and DataRobot

• When automation helps and when it just generates prettier mistakes

• Dev confessionals: “The time AutoML destroyed our budget in 8 hours”

Episode 9: “OpenAI, Meta, Mistral, and the Model Wars Deathmatch”

• The LLM leaderboard is rigged and everyone’s cooking their own benchmarks

• Closed vs. open source: the real stakes behind the GitHub stars

• Drafting the ultimate AI model team like it's fantasy football

Episode 10: “Fine-Tuning is Like Botox: Used Right, It’s Magic. Used Wrong, You Look Melted.”

• When fine-tuning makes sense, and when you should just stop

• LoRA, embeddings, vector stores, and why nobody documents anything

Season Ending: What if we just used less AI?

• Is there an app for that?