You’re Using AI Wrong: Build the System, Not Just the Prompt /w Lexi Pasi episode artwork

EPISODE · May 8, 2026 · 43 MIN

You’re Using AI Wrong: Build the System, Not Just the Prompt /w Lexi Pasi

from Domesticating AI · host SoyPete Tech

Recorded: April 14, 2026Most people using AI today are still users.They open ChatGPT, call an API, and get an answer.And honestly… it works.But that’s not the same as building with AI.In this episode of Domesticating AI, we break down the difference between AI users and AI practitioners—and why that shift matters if you want reliable systems.We’re joined by Alexandra “Lexi” Pasi, PhD, CEO of Lucidity Sciences, to talk about what it actually means to own the system around AI:why calling an API is still user behaviorwhat changes when you build the harnesshow agent systems actually fail (loops, cost, drift)why switching models isn’t a reliability strategyhow to add layers—constraints, validation, and control flowwhy engineering discipline matters more with AI, not lessIf you’ve built your first AI agent, workflow, or coding loop—this is the “now what?” episode.Alexandra Pasi is the CEO of Lucidity Sciences, where she works at the intersection of mathematics, machine learning, and real-world system design.She holds a PhD in Mathematics from Baylor University and specializes in building analytical and algorithmic systems that bring structure to complex, uncertain environments.🔗 LinkedIn: https://www.linkedin.com/in/alexandrapasi/🔗 Lucidity Sciences: https://luciditysciences.comGoogle TurboQuant (LLM compression research)https://research.google/blog/turboquant-redefining-ai-efficiency-with-extreme-compression/Anthropic Claude Mythos Preview (security-focused model)https://red.anthropic.com/2026/mythos-preview/Project Glasswing (Anthropic security initiative)https://www.anthropic.com/glasswingKarpathy Autoresearch (self-improving training loop)https://github.com/karpathy/autoresearchKitaru (durable agent execution framework)https://github.com/zenml-io/kitaruSubscribe on YouTubeFollow on Spotify & Apple PodcastsSupport the show on Patreon:👉 https://patreon.com/DomesticatingAIPodcastKeep your AI on a leash.🧾 Episode Summary👤 Guest: Alexandra “Lexi” Pasi, PhD🔗 Topics & Links Mentioned🔔 Follow & Support

Episode metadata supplied by the publisher feed · Published May 8, 2026

Embed this episode

Ready to play

You’re Using AI Wrong: Build the System, Not Just the Prompt /w Lexi Pasi

0:00 43:25

No transcript for this episode yet

We transcribe on demand. Request one and we'll notify you when it's ready — usually under 10 minutes.

No similar episodes found.

Frequently Asked Questions

How long is this episode of Domesticating AI?

This episode is 43 minutes long.

When was this Domesticating AI episode published?

This episode was published on May 8, 2026.

Can I download this Domesticating AI episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.
URL copied to clipboard!