<p><em>> It’s the agent writing the code. And it’s the development loop of writing the code, building testing, write the code, build test and iterating. And so</em><strong><em> I do think we’ll see for many types of software, a shift away from Python towards other programming languages</em></strong><em>. I think Go is probably the best language for those like other types of software projects. And like I said, I haven’t written a line of Go code in my life.</em></p><p><em>– Wes McKinney </em>(<strong>creator</strong> of <strong>pandas</strong> <strong>Principal Architect</strong> at <strong>Posit)</strong>,</p><p><strong>Wes McKinney</strong>, <strong>Marcel Kornacker</strong>, and <strong>Alison Hill</strong> join Hugo to talk about the architectural shift for multimodal AI, the rise of “agent ergonomics,” and the evolving role of developers in an AI-generated future.</p><p><strong>We Discuss:</strong></p><p>* <strong>Agent Ergonomics:</strong> Optimize for <strong>agent iteration speed</strong>, shifting from human coding to fast test environments, potentially favoring languages like <strong>Go</strong>;</p><p>* <strong>Adversarial Code Review:</strong> Deploy <strong>diverse AI models</strong> to <strong>peer-review agent-generated code</strong>, catching subtle bugs humans miss;</p><p>* <strong>Multimodal Data Verbs:</strong> Make operations like <strong>resizing</strong> and <strong>rotating</strong> native to your database to eliminate <strong>data-plumbing bottlenecks</strong>;</p><p>* <strong>Taste as Differentiator:</strong> Value <strong>“taste”</strong>—the ability to <strong>curate</strong> and refine the best output from countless <strong>AI-generated options</strong>—over sheer execution speed;</p><p>* <strong>100x Software Volume:</strong> Embrace <strong>ephemeral, just-in-time software</strong>; prioritize aggressive generation and <strong>adversarial testing</strong> over careful planning for quality.</p><p>You can also find the full episode on <a target="_blank" href="https://open.spotify.com/show/3yuz89gqAhcMcdy3SZPe4X?si=AKl2jvIARD2Liw1bBH2Nng&#38;nd=1&#38;dlsi=d536dc4dceca4619">Spotify</a>, <a target="_blank" href="https://podcasts.apple.com/us/podcast/vanishing-gradients/id1610318868">Apple Podcasts</a>, and <a target="_blank" href="https://youtube.com/live/UwdpNxHZDwI?feature=share">YouTube</a>.</p><p><a target="_blank" href="https://notebooklm.google.com/notebook/6225d2fa-6a36-4a48-a0ab-131736c1652e">You can also interact directly with the transcript of the workshop & fireside chat here in NotebookLM</a>: If you do so, let us know anything you find in the comments!</p><p>👉 <strong><em>Want to learn more about Building AI-Powered Software? Check out our </em></strong><a target="_blank" href="https://maven.com/hugo-stefan/building-ai-apps-ds-and-swe-from-first-principles?promoCode=vgfs"><strong><em>Building AI Applications course</em></strong></a>. It’s a live cohort with hands on exercises and office hours. Here is a <a target="_blank" href="https://maven.com/hugo-stefan/building-ai-apps-ds-and-swe-from-first-principles?promoCode=vgfs">discount code</a> for readers. 👈</p><p>This was a fireside chat at the end of a livestreamed workshop we did on building multimodal AI systems with Pixeltable. Check out the full workshop below (all code here on Github):</p><p><strong>Links and Resources</strong></p><p>* <a target="_blank" href="https://www.linkedin.com/in/wesmckinn/">Wes McKinney on LinkedIn</a></p><p>* <a target="_blank" href="https://www.linkedin.com/in/marcelkornacker/">Marcel Kornacker on LinkedIn</a></p><p>* <a target="_blank" href="https://www.linkedin.com/in/apreshill/">Alison Hill on LinkedIn</a></p><p>* <a target="_blank" href="https://spicytakes.org">Spicy Takes</a></p><p>* <a target="_blank" href="https://allisonhorst.github.io/palmerpenguins/">Palmer Penguins</a></p><p>* <a target="_blank" href="https://pixeltable.com/">Pixeltable</a></p><p>* <a target="_blank" href="https://posit.co/">Posit</a></p><p>* <a target="_blank" href="https://positron.posit.co/">Positron</a></p><p>* <a target="_blank" href="https://github.com/apreshill/pxt-primetime">Building Multimodal AI Systems Workshop Repository</a></p><p>* <a target="_blank" href="https://docs.pixeltable.com/howto/cookbooks/agents/llm-tool-calling#using-mcp-servers-as-tools">Pixeltable Docs: LLM Tool Calling with MCP Servers</a></p><p>* <a target="_blank" href="https://docs.pixeltable.com/howto/providers/working-with-pydantic">Pixeltable Docs: Working with Pydantic</a></p><p>* <a target="_blank" href="https://lu.ma/vanishinggradients">Upcoming Events on Luma</a></p><p>* <a target="_blank" href="https://www.youtube.com/@vanishinggradients">Vanishing Gradients on YouTube</a></p><p>* <a target="_blank" href="https://youtube.com/live/UwdpNxHZDwI?feature=share">Watch the podcast video on YouTube</a></p><p>* <a target="_blank" href="https://maven.com/hugo-stefan/building-ai-apps-ds-and-swe-from-first-principles?promoCode=vgfs">Join the final cohort of our Building AI Applications course in March, 2026 (25% off for listeners)</a></p><p>https://maven.com/hugo-stefan/building-ai-apps-ds-and-swe-from-first-principles?promoCode=vgfs</p><p></p><p><strong>What people said during the workshop</strong></p><p>“I think the interface looks amazing/simple. Strong work! 🦾” — @goldentribe</p><p></p><p>“This is quite amazing. Watching this I felt the same way when I first leant pandas, NumPy and scikit and how well i was able to manipulate and wrangle data. PixelTable feels seamless and looks as good as those legendary frameworks but for Multimodal Data.” — @vinod7</p><p></p><p>“This is all extremely cool to see, I love the API and the approach.” — @steveb4191</p><p></p><p>“Thanks so much, Hugo! That was very insightful! Great work Alison and Marcel!” — @vinod7</p><p></p><p>“Just wrapped up watching a replay of the Pixeltable workshop. So cool!! Love the notebooks and working examples. The important parts were covered and worked beautifully 🕺” — @therobbrennan</p><p></p><p>👉 <strong><em>Want to learn more about Building AI-Powered Software? Check out our </em></strong><a target="_blank" href="https://maven.com/hugo-stefan/building-ai-apps-ds-and-swe-from-first-principles?promoCode=vgfs"><strong><em>Building AI Applications course</em></strong></a>. It’s a live cohort with hands on exercises and office hours. Here is a <a target="_blank" href="https://maven.com/hugo-stefan/building-ai-apps-ds-and-swe-from-first-principles?promoCode=vgfs">discount code</a> for readers. 👈</p> <br/><br/>Get full access to Vanishing Gradients at <a href="https://hugobowne.substack.com/subscribe?utm_medium=podcast&#38;utm_campaign=CTA_4">hugobowne.substack.com/subscribe</a>